Understanding the structure and processes of primary health care for young indigenous children

KEYWORDS: Indigenous health; health services; health systems; paediatrics; epidemiology

Internationally, the health and social wellbeing of young indigenous children are of major concern.¹ In Australia, young Aboriginal and Torres Strait Islander children (hereafter ‘indigenous’) remain a high-risk group for experiencing adverse health and social outcomes such as otitis media,² child neurodevelopment delay³ and birth outcomes such as prematurity and low birthweight⁴ compared to non-indigenous Australian children. Despite this, improvements in primary health care, coupled with major policy and funding changes, has resulted in an increase of important child health indicators including child health assessments and vaccination coverage.^4–7

Primary health care plays an important role in the delivery of community and preventive health services. However, providing high-quality care remains an ongoing challenge.⁸ Detailed measurement and evaluation of the quality of care delivered to indigenous children is needed to track and improve service delivery. This can be determined through understanding the relationship and interplay between the three categories of quality of care: structure (attributes and organisational structures that define a health system); processes (delivering and receiving care); and outcomes (the consequences or effect of care on health status).⁹ Good structural systems are expected to lead to good processes of care and ultimately improved outcomes.⁹ It is therefore important to objectively assess the relationship between these three categories and service delivery to children in real world situations.

The Assessment of Chronic Illness Care (ACIC) tool was developed to help health services understand the organisation of care within their systems, identify areas for improvement and evaluate the level and nature of these changes for people living with a chronic disease.¹⁰ The ACIC team identified six areas of system change: delivery system design, self-management, clinical information systems, linkages to community resources, decision support, and organisation of the health system.¹⁰ In 2005, the Audit and Best Practice for Chronic Disease (ABCD) programme (a continuous quality improvement (CQI) programme in Australia) modified the ACIC tool, added three items (cultural competence, laboratory management and pharmacy management) and developed the Systems Assessment Tool (SAT).¹¹ The SAT has enabled indigenous health services to assess their health-care systems and improve the quality of care they provide.^11,12

To date, the SAT has been used to assess the quality of care for diabetes and pregnancy.^11,13,14 The SAT has yet to be used to assess on a broad scale the quality of care delivered through organisations of care (structures) for indigenous children and key process of care indicators (PoCIs) for important childhood health and social issues, in particular, social and emotional wellbeing, anaemia and child neurodevelopment. Therefore, the objectives of this study were to determine whether there was an association between social and emotional wellbeing, anaemia and child neurodevelopment PoCIs and: (i) primary health-care service and child characteristics; and (ii) organisational health service structures. It was hypothesised that fully supported organisations and structures within health services would result in increased improvement in processes of care for indigenous children.

This was a retrospective cross-sectional study of 1554 child health audits that included SAT data from remote, rural and urban primary health-care services that participated in the ABCD programme in Queensland, Northern Territory, South Australia and Western Australia from 2012 to 2014. The most recent record for each child was included.

Annual child health audits from participating primary health-care services were completed by primary care staff who had received training by ABCD educators.^15,16 Files had to meet the following criteria to be eligible for auditing: (1) child is aged 3 months to 14 years at the audit date; (2) child is a resident in the community for at least 6 months (or half of the time since birth if aged <6 months); and (3) child has no major health anomalies such as heart defects or inherited disorders.

A random sample of at least 30 files was selected for audit from each participating primary health-care service. The sampling process included stratification of sex to ensure similar numbers. The auditors read each client file (electronic and paper) and recorded information in a standardised pre-coded data collection tool. Child characteristics included: date of birth, age, sex, indigenous status, attendance at the primary care centre in the previous 12 months, reason for the last attendance (acute care, health check, vaccination, other) and receipt of any child health checks in the last 12 months (Australian Commonwealth funded [Medicare 715] or other child health check). Health centre characteristics included governance (Aboriginal community-controlled health service or government operated), location (urban, rural or remote) and number of CQI audits the primary care centre had completed. The auditors scored ‘yes’ in the audit tool if there had been any documentation in the client file in the last 12 months, ‘no’ if there was no documentation and ‘not applicable’ if a service was not recommended or scheduled within that jurisdiction.

The ABCD audit tool included 11 pre-coded items about social and emotional wellbeing services, seven on anaemia and six on child neurodevelopment. Descriptions of these items are provided in Supplementary material table 1 (available at journal’s website).

Table 1. Key characteristics of audits and health-care centres for indigenous children aged 3–59 months

The SAT (Supplementary material table 2) included five components (delivery system design; information systems and decision support; self-management support; links with community, other health service and other services; and organisational influence and integration), with multiple items for each component. The SAT was self-completed in each primary health-care service by staff and a trained CQI facilitator. Each item within a component was scored as 0–2 (limited), 3–5 (basic), 6–8 (good) and 9–11 (fully developed) by the health service. The tool included a brief description of each component and item to help health service staff decide on their existing support. The CQI facilitator helps health service staff reach agreement about what best represents their health system. Each component score was calculated as the mean of the individual items. The overall organisation of the health system score was the mean of the four component scores. We did not include the self-management component of the SAT due to perceptions that this was of limited relevance to the study outcomes.

Table 2. Associations between key characteristics and process of care indicators in indigenous children aged 3–59^† months

We defined PoCIs as:

• Social and emotional wellbeing : using four items comprising advice provided to parents or carers at least once in the last 12 months about: domestic environment, social support, housing condition and child stimulation for all children aged 3–59 months.

• Anaemia : using two items comprising advice provided at least once in the last 12 months about nutrition and haemoglobin, documented in the last 12 months for all children aged 6–59 months.

• Child neurodevelopment : using five items comprising assessment provided in the last 12 months for parent–child interaction (<2 years), developmental milestones, vision and hearing testing for all children aged 3–59 months. Advice about physical and mental stimulation of the child was also included for all children aged 3–59 months.

The PoCIs were developed using the Primary Clinical Care Manual for Queensland,^17,18 the Central Australian Rural Practitioners Association (CARPA) standard treatment manuals for Northern Territory and South Australia¹⁹ and the Kimberley Aboriginal Medical Service guidelines.^20,21 The Medicare Benefits Schedule (MBS) child health check²² and the National guide to a preventive health assessment for Aboriginal and Torres Strait Islander people²³ were also consulted in the development of the PoCIs. To develop the PoCIs, there had to be commonality in individual items between jurisdictions. The PoCIs were dichotomised into a score of ‘yes’ if an audit record showed evidence that all items had been completed or ‘no’ if records were partially or not completed (Supplementary material table 3).

Table 3. Association between delivery systems design and process of care indicators in indigenous children aged 3–59^† months

Descriptive statistics were calculated as counts and percentages for all categorical data, and median and interquartile ranges (IQR, 75% percentile – 25% percentile) for continuous data. Data analyses were conducted using STATA 13.1 (StataCorp, College Station, TX, USA).

To examine the effect of primary health-care service and child characteristics on the probability of having a PoCI, multilevel binomial models with an exchangeable correlation structure and robust standard errors were used. Adjusted logistic regression models were fitted using generalised estimating equations and the primary health-care service as the clustering variable. Odds ratios and 95% confidence intervals (95% CI) were derived. Important explanatory variables were constructed a priori and included: sex, year of data collection, geographic location, governance and CQI participation.

To assess associations between SAT components and the three PoCIs, crude and adjusted logistic regression models were fitted using generalised estimating equations and the primary health-care service as the clustering variable. Multilevel binomial models with an exchangeable correlation structure and robust standard errors were also constructed, and odds ratios and 95% CIs were derived. Important explanatory variables were constructed a priori and included: year of data collection, geographical location, governance, CQI participation and the number of health areas SAT was related to.

Ethics approval was obtained from all Human Research Ethics Committees in the states and territories involved: the Human Research Ethics Committee of the Northern Territory Department of Health and Menzies School of Health Research (HREC-205 EC00153); Central Australian Human Research Ethics Committee (HREC-12–53); Queensland Human Research Ethics Committee of the Darling Downs Health Services District (HREC/11/QTDD/47); South Australian Indigenous Health Research Ethics Committee (04–10–319); Curtin University Human Research Ethics Committee (HR140/2008); Western Australian Country Health Services Research Ethics Committee (2011/27); Western Australian Aboriginal Health Ethics Committee (111–8/05); University of Western Australia Human Research Ethics Committee (RA/4/1/5051); and the Australian National University (2017/560).

During 2012–14, there were 1554 records audited from 74 primary health-care services that completed the SAT (Table 1). Most health services (94.6%, 70/74) were government run, serviced a population of <500 people (58.1%, 43/74), were in remote locations (87.8%, 65/74) and had participated in three or more CQI cycles (73.0%, 54/74) (Table 1). Approximately half (54.6%, 848/1554) of records audited were for children aged between 24 and 59 months.

Less than one-third of records (32.0%, 449) had a social and emotional wellbeing PoCI, 56.6% (791) had an anaemia PoCI and just under half (49.3%, 430) had a child neurodevelopment PoCI (Table 2). Children aged 12–23 months had increased odds of receiving a PoCI for social and emotional wellbeing (aOR 1.35, 95% CI 1.01–1.49), anaemia (aOR 1.68, 95% CI 1.30–2.18) and child neurodevelopment (aOR 1.80, 95% CI 1.12–2.90) compared to children aged 24–59 months (Table 2). Children who received acute care were less likely to have a PoCI for social and emotional wellbeing (aOR 0.74, 95% CI 0.57–0.96), anaemia (aOR 0.62, 95% CI 0.49–0.77) and child neurodevelopment (aOR 0.61, 95% CI 0.44–0.85) than children who received a child health check (Table 2).

Two items within the delivery system design SAT component were significantly associated with the anaemia PoCI. For every one point increase in the team structure and function item, there was a 14% greater odds of having an anaemia PoCI (aOR 1.14, 95% CI 1.01–1.27) (Table 3). A similar trend was shown for care planning where for every one point increase in the care planning item, there was also a 14% greater odds of having the anaemia PoCI (aOR1.14, 95% CI 1.01–1.29) (Table 3). Social and emotional wellbeing and child neurodevelopment PoCIs were not influenced by delivery systems design (Table 3). There was no association between information systems and decision support, links with community and other health and non-health services or organisational influence and integration on any of the three PoCIs (Table 4–6). For all three PoCIs, there was little difference in the median and interquartile ranges for each system assessment item and component between children receiving the PoCI and children who did not (Supplementary material table 4).

Table 4. Association between information systems and decision support and process of care indicators in indigenous children aged 3–59^† months

Table 5. Association between links with community, other health services and other services and process of care indicators in indigenous children aged 3–59^† months

Table 6. Association between organisational influence and integration and process of care indicators in indigenous children aged 3–59^† months

There was wide variation on the delivery of PoCIs, with 32.0% of children receiving PoCIs for social and emotional wellbeing, 56.6% for anaemia and 49.3% for child neurodevelopment. Children who were aged 12–23 months were more likely to receive the PoCIs than children aged 24–59 months. Contrary to our expectations, there was little association between an organisation’s health structures and processes of care for social and emotional wellbeing and child neurodevelopment. However, we found that the delivery system design component, which included items for team structure and function and care planning, were associated with the process of care provided for anaemia.

Based on the reporting of child health indicators, CQI has improved the delivery of many child health milestones and brief interventions over time.⁴ As a result, we anticipated that 50% of records would achieve our PoCIs. However, social and emotional wellbeing (32%) fell well short of this. In contrast, anaemia (60%) and child neurodevelopment (49%) reached the expected target. Our data also indicate that the provision of care varies greatly, depending on the routine service provided.

There has been much concern in primary health-care centres about the high levels of iron-deficiency anaemia in young indigenous children.²⁴ Substantial emphasis in recent years on improving these rates in remote regions has included improving primary and secondary prevention through implementing CQI initiatives, health sector forums and community engagement. In our study, elements of team structure and functioning including team leadership, defining roles and responsibilities and building capacity, as well as care planning that includes planning as part of routine practice, and consistency with best practice guidelines, were shown to be positive in improving anaemia care. Capacity building with health service providers for anaemia care has fostered important changes in health services and provided quality care to children.²⁵ Alternatively, non-adherence to guidelines has resulted in poor management of children with anaemia.²⁶

In contrast, there has been little focus on the more complex processes of care needed for social and emotional wellbeing and child neurodevelopment, and this is reflected in the lower PoCIs that we reported for these two conditions.²⁷ The delivery of child neurodevelopment assessments has been shown to vary across primary health-care services, and researchers have recently called for a more system-wide approach to improve delivery, recording and monitoring.²⁸ There has been a recent emphasis on the importance of child neurodevelopment and the delivery of social and emotional services in family-centred care practice. Thus, it is expected that we will see improvements in the provision of care for these important areas in the coming years.^29,30

It is our understanding that this is the first study to investigate associations between a standardised SAT assessment of quality of care and the processes of care delivered to indigenous children in primary health care. It has previously been shown that a health-care system’s organisational influence and integration is positively associated with the quality of care provided to adults for diabetes control (HbA1C), blood pressure and total cholesterol levels.³¹ However, we found no association between the organisation of health systems and the provision of care for children as measured by our neurodevelopment and social and emotional wellbeing PoCIs. Almost 70% of the health services had completed at least three or more SAT cycles, thus this may have improved the organisation of all the health-care systems in terms of these PoCIs over time. It is also possible that other factors have resulted in this lack of association, including lack of assessment of communication and patient-centred care, which were not included in our analysis and the potential for under-reporting in health records. Despite this, processes of care and organisational systems in primary care services were shown to be important for the optimal management of anaemia in indigenous children and should be enhanced.

There are several limitations to this study. Although guidance and facilitation was provided to local health centre staff and managers to complete the SAT, in practice, they were largely completed by front-line primary health-care teams without direct standardised support. This is likely to influence how the tool was completed. It was decided that the process of care delivered should be maximal and therefore included all elements documented. Although the PoCIs have not been validated, we believe this study has demonstrated an important use of them. In addition, we constructed them through ensuring that they were specific, measurable, attainable, relevant and trackable. It is also possible that for some levels of care, there was no or little documentation of this in the health-care records.

This was a cross-sectional study so we could only report associations and could not assess causality. The positive results seen in our analyses may be the result of type 1 error; however, given that our P values were not borderline and we have narrow confidence intervals, we are confident in our results. Due to the voluntary nature of participation by primary health-care services in this study, the findings are not necessarily generalisable to all primary health-care services. Most of the health services were government-run (94.6%), located in remote areas (87.8%) and serviced populations of <500 (58.1%) people. This also limits the potential generalisability of findings to other health services, in particular Aboriginal Community-Controlled Health Organisations.

Our study found that organisational health service structures, which included items for team structure and function and care planning, were associated with quality of anaemia care. This study provides evidence that organisation of health services is associated with the prevention and management of anaemia for young indigenous children. In addition, our young indigenous children aged 24–59 months are not receiving care for important social and health indicators. Child health checks are an important avenue to ensuring quality care is provided.

This research was supported by an Australian Government Research Training Program (RTP) Scholarship. This study was also funded by the National Health and Medical Research Council of Australia. The funders had no role in the design of the study and collection, analysis and interpretation of data, and in writing the manuscript.

The authors declare that they have no competing interests.