Entry into long-term care disrupts primary care continuity—but only at entry: a national cohort study of transitional and post-entry continuity in facility- and home-based care in Australia

Abstract

Background

Continuity of primary care is central to older adults’ health and wellbeing, but entry into long-term care can disrupt established care relationships.

Objectives

This study aimed to characterize how continuity of care evolves across the transition into long-term care and the post-entry period, comparing facility-based and home-based care, with analyses stratified by metropolitan and non-metropolitan location.

Methods

A national cohort study of Australians aged ≥65 who entered long-term care in 2019 and had at least three primary care visits within 6 months was conducted. Transitional continuity was defined by whether individuals continued seeing their pre-entry provider. Post-entry continuity was measured using the Usual Provider Index. Cohort differences were examined using regression models.

Results

Among 69 081 entrants (home-based = 27 915; facility-based = 41 166), 72% of those entering facilities changed to a new provider, compared with 10% in home-based long-term care, p < .001. In facility-based long-term care, switching to a new provider was more common in metropolitan areas than in non-metropolitan areas. Post-entry continuity was high in both settings, with 78%–80% of visits delivered by a usual provider.

Conclusions

Entry into facility-based long-term care substantially disrupts established primary care relationships, but continuity is rapidly re-established after entry. The greater disruption observed in metropolitan areas suggests that factors beyond relocation, such as primary care availability and delivery models, shape transitional continuity. These findings identify long-term care entry as a key intervention point while indicating that post-entry care models effectively consolidate continuity.

Key messages.

• Entry into long-term care facilities disrupted established primary care ties.

• Seventy-two percent of facility entrants changed to a new provider.

• Provider switching was more common in metropolitan than non-metropolitan areas.

• Post-entry continuity was high, with 78%–80% of visits to a usual provider.

• Structured handover may help manage discontinuity at facility entry.

Background

Globally, 10%–20% of adults over 65 years of age receive long-term care (LTC) in publicly subsidized aged care programs [1], and more than 1.5 million people receive government-subsidized LTC services in Australia every year [2]. Entry into LTC is a vulnerable period for older adults. Often precipitated by increasing frailty and care needs, entry is a major life transition that can unsettle social and living arrangements and disrupt established healthcare relationships [3]. While most older adults in the community maintain ongoing relationships with a usual primary care provider, covering roughly half to three-quarters of attendances [4, 5], this relational continuity can drop sharply around entry into facility-based LTC due to relocation and associated provider changes [6,7,8,9]. However, it is unclear if such transitional disruptions extend to low post-entry relational continuity, or if new entrants receive continuous primary care from a main provider once established in LTC. Furthermore, variation in continuity across factors such as location in a metropolitan versus non-metropolitan area or between for-profit, not-for-profit, and government-run providers is unknown. This is especially relevant in the context of significant efforts to facilitate continuous and person-centred primary care delivery in LTC [10,11,12].

Continuity disruptions are clinically consequential. In systems like Australia's, where the general practitioner (GP) is the first point of contact and central to integrated care, continuity disruptions can affect care quality. Among community-dwelling older adults, continuity of primary healthcare has been associated with lower risks of mortality [13] and hospitalizations [14, 15], greater patient satisfaction [16], and lower healthcare system costs [16]. Similarly, a recent systematic review reported that higher relational continuity post care-entry reduces hospitalizations and emergency department visits among older adults receiving LTC at home [3], and population-based cohort studies in home and facility settings have linked transitional and post-entry continuity to lower risks of mortality and hospitalization [17, 18]. Due to these substantial benefits, ensuring relational continuity across the transitional entry period and post-entry is a clear priority across both home-based and facility-based LTC.

Securing transitional continuity with a preexisting community GP is challenging in facility settings often due to relocation and limited GP availability. When entering facility-based LTC, 75% of people move away from their immediate area, with an average relocation distance of 54 km, and for the 1.9% of Australians living in remote areas more than 400 km [19]. Even when limited relocation distance would allow for continuity, only about 10% of practicing GPs reported working in LTC facilities in 2024 [20], and more than one-third of those have indicated plans to reduce or cease this work [21], reflecting recognized remuneration and workflow disincentives [21,22,23,24]. This suggests that transitional continuity may be especially low in non-metropolitan areas, with higher relocation distances and lower GP availability.

The limited evidence on transitional continuity from Ontario, Canada [8], and people living with dementia in Australia [9], suggests that only 12%–29% retain their pre-entry GP when entering facility-based LTC. However, it remains unclear whether this low level of transitional continuity applies to the broader population. This question is relevant in the Australian context, where there is no consistent model for managing informational continuity during individuals’ transitions into facility-based LTC. In contrast to countries such as the Netherlands, where GPs formally transfer care to nursing home physicians and systematically hand over patient-relevant information at LTC entry [25], disruptions in transitional continuity in Australia are likely to be accompanied by losses of informational continuity.

Importantly, if transitional continuity is challenging to achieve in facility-based LTC, and there is no structured handover process ensuring informational continuity, post-entry continuity becomes even more decisive, as it may have to compensate for potential detrimental impacts of transitional disruptions. Recognizing the risks of continuity disruptions, especially around entry into LTC facilities, providers and policymakers employ a variety of facility-focused primary care delivery models, depending on facility context and GP availability. Reported models include (1) continuity models where community GPs continue care post-entry, (2) panel or facility-linked arrangements, and (3) team-based or in-reach models involving specialized or hospital-affiliated providers [12]. While the continuity model aspires to secure transitional continuity (pre- to post-entry), others are focused on providing continuous and team-based care post-entry [26]. Exact prevalences of these models are unclear, but most GPs serving LTC facilities operate from community-based practices, and facility-based physicians, as often seen in the United States, are less frequent [27]. However, despite this diversity of care models designed to support person-centred and continuous primary care post-entry, there is no population-based evidence quantifying the level of post-entry continuity achieved across these models.

The aim of this study was to provide a comprehensive population-based investigation into the level of continuity of care experienced by older people entering LTC at home or in facilities. The study aimed to (1) assess whether LTC entrants change their usual primary healthcare provider across the entry transition period, (2) quantify continuity in the 180 days post-entry, and (3) describe differences in transitional and post-entry continuity by LTC setting (home-based versus facility-based), location, and ownership type. For these aims, LTC entry was defined as the commencement of a publicly subsidized home-based or facility-based LTC service. LTC at home was included as a comparator because ageing-in-place is expanding [28], residential relocation is minimal, and—given the similar health status of recipients [29]—it provides a relevant benchmark against which to interpret facility-based outcomes.

Methods

Study design, setting and data sources

A population-based retrospective cohort study using data from the Registry of Senior Australians (ROSA) National Historical Cohort [30] was conducted. The National Historical Cohort has been described in detail elsewhere [30]. At the time of analysis, the cohort included a large, linked collection of national and state-based health, LTC and social service use data collections for approximately 3.5 million Australians who accessed LTC between 2002 and 2020. For this analysis, we drew on multiple ROSA data collections, including LTC eligibility assessments for facility- and home-based LTC services (including respite and transition care), service episode records, LTC entry needs assessments, Pharmaceutical Benefits Scheme dispensing claims, Medicare Benefits Schedule subsidized service claims, and death registrations from the National Death Index.

Study cohort

National entrants into LTC in facilities (N = 41 166) and at home (N = 27 915) in 2019, aged ≥65 years, who did not identify as Aboriginal or Torres Strait Islander, and were not Department of Veterans’ Affairs card holders were studied. To ensure that continuity estimates were computable in line with established methods [9, 31], this cohort was restricted to individuals who remained in care for at least 180 days and had at least three primary healthcare attendances during this period. The year of 2019 was selected as the study period because it represented the most recent period not affected by the COVID-19 pandemic available at the time of analysis. LTC is federally subsidized in Australia. During the study period, LTC at home was funded by the Australian Government Home Care Package Programme [32] and LTC in facilities was funded by the Residential Care Programme [33]. Department of Veterans’ Affairs card holders represent 11% of those in LTC in facilities and 4% of those accessing LTC at home and were not included due to their concessional entitlements that may lead to a differential level of healthcare service access. Individuals of Aboriginal and Torres Strait Islander background represent approximately 2% of new LTC admissions in each year and analyses using their data require leadership from Aboriginal and Torres Strait Islander people and specific Indigenous governance and ethics approvals, which were not part of our current study.

Continuity of care

Continuity of care was based on individuals’ primary healthcare attendances. The most common primary healthcare attendance in Australia is a general attendance provided by a GP in the community, but to cover different models of primary healthcare delivery in LTC at home or in facilities, other attendances were included in the ascertainment of continuity of care (e.g. after-hours attendances, on-site and in-home attendances, and primary healthcare attendances provided by medical practitioners other than GPs). No encounters with specialists or hospitals were included in the ascertainment (see Supplementary Table S1 for coding and encounters included).

Transitional continuity was determined by comparing the most frequently seen primary healthcare provider within the first 180 days after entry to the providers seen in the 24 months preceding entry, aligned with established methods [9]. Transitional continuity was classified as new provider (i.e. most frequently seen provider after entry was never seen prior to entry), known provider (i.e. most frequently seen provider after entry was seen at least once prior to entry), and usual provider (i.e. most frequently seen provider did not change from before to after entry) as previously defined [9] (Supplementary Figure S1). Post-entry continuity in the first 180 days was measured using the Usual Provider Index (UPI) [31]. The UPI is defined as the proportion of attendances provided by the most frequently seen primary healthcare provider.

Covariates

Individuals’ age (categorized into 65–69, 70–74, 75–79, 80–84, 85–89, and ≥90 years old), sex, and number of health conditions (0–2, 3–5, 6–10, and ≥11) estimated using the RxRisk-V medication-based comorbidity score measure [34] with a 180-day look-back period at care entry, location (metropolitan versus non-metropolitan, based on the Accessibility/Remoteness Index of Australia [35]), and ownership type (for-profit, not-for-profit, and government-run) were used for risk adjustment. Dementia at entry into care was ascertained using health conditions reported at the facility-based LTC entry assessments, the LTC eligibility assessments, and RxRisk-V derived from medication dispensing (6-month look-back).

Analysis

Descriptive statistics characterized the individuals examined. Crude and risk-adjusted prevalences of both continuity measures were estimated, with crude estimates additionally stratified by location and ownership type. To calculate risk-adjusted prevalences and test differences across LTC settings, we pooled the cohorts and fitted logistic regression models with cohort membership (LTC in facilities or at home) as the key factor, adjusting for age, sex, comorbidity score, location, and ownership type. Transitional continuity, classified as usual, known, or new provider, was analysed using multinomial logistic regression. Post-entry continuity was analysed using a grouped binomial logistic regression, using the number of attendances with the usual provider as the numerator and the total number of post-entry primary healthcare attendances as the denominator. There was minimal missing data on the examined variables (0.2%, 118/69 081), and a complete case analysis was conducted. We used cluster-robust sandwich standard errors to account for repeated observations on individuals who transitioned from home-based to facility-based LTC. Average marginal effects were then calculated on the probability scale to yield adjusted prevalences per cohort and the absolute probability difference between them. A sensitivity analysis was conducted to examine whether the findings differed between those with and without a diagnosis of dementia. To assess the impact of restricting the cohort to people who remained in care for at least 180 days and had at least three primary care attendances, a sensitivity analysis examined transitional continuity among all older adults who entered facility-based LTC in 2019 with at least one post-entry attendance. At the first post-entry visit, the provider was classified as new, previously seen, or the person′s usual pre-entry provider, with analyses stratified by days in care (≤180 days, >180 days).

All analyses were conducted using R v4.3.0 (R Core Team, Vienna, Austria).

Results

Cohort

The facility-based LTC cohort (n = 41 166) had a median age of 85 years (interquartile range [IQR] = 79–89), a median number of five (IQR = 3–7) health conditions, a median number of 16 (IQR = 10–23) primary care attendances in the 180 days post-entry, and 61.2% (n = 25 212) were female (Table 1). The home-based LTC cohort (n = 27 915) had a median age of 82 years (IQR = 77–87), a median number of five (IQR = 4–7) health conditions, a median number of nine (IQR = 6–14) primary care attendances in the 180 days post-entry, and 62.8% (n = 17 518) were female (Table 1). The majority of both cohorts were located in metropolitan areas (facility-based: 70.5%, home-based: 72.4%) and received care from not-for-profit providers (facility-based: 54.2%, home-based: 65%) (Table 1).

Table 1.

Study characteristics.

Variable	Categories	Facility-based LTC	Home-based LTC
Total N	N	41 166	27 915
Sex, n (%)	Female	25 212 (61.2)	17 518 (62.8)
	Male	15 954 (38.8)	10 397 (37.2)
Median age at entry into LTC [Q1, Q3]		85 [79, 89]	82 [77, 87]
Median number of health conditionsa [Q1, Q3]		5 [3, 7]	5 [4, 7]
Dementia, n (%)		21 161 (51.4)	4024 (14.4)
Median primary care attendances in 180 days post-LTC entry [Q1, Q3]		16 [10, 23]	9 [6, 14]
Location, n (%)	Metropolitan	29 015 (70.5)	20 221 (72.4)
	Non-metropolitan	12 039 (29.2)	7692 (27.6)
	Missing	112 (0.3)	2 (0.1)
Ownership type, n (%)	Government	1530 (3.7)	1969 (7.1)
	Not-for-profit	22 306 (54.2)	18 136 (65.0)
	For-profit	17 330 (42.1)	7810 (28.0)

Abbreviation: Q, quartile.

^aRxRisk-V comorbidity score.

Transitional continuity

Compared with their usual provider during the 2 years before entry, 71.5% (29 451/41 166) of those who entered facility-based LTC changed to a new provider after entry (versus 10.2%, 2849/27 915, of those who entered home-based LTC), 11.7% (4814/41 166) saw a known provider (57.8%, 16 145/27 915, home-based), and 16.8% (6901/41 166) retained their usual provider (32.0%, 8921/27 915, home-based) (Table 2). After adjustment, those entering facility-based LTC had a 61.8-percentage-point higher probability of changing to a new provider than those entering home-based LTC [adjusted probabilities 71.9% versus 10.1%, difference = 0.618; 95% confidence interval (CI) 0.605–0.629, p < 0.001] (Table 2).

Table 2.

Crude and risk-adjusted primary healthcare continuity estimates for individuals in long-term care in facilities or at home, and by service location and ownership type.

Measure	Transitional continuity			Post-entry continuity (UPI)
	New provider	Known provider	Usual provider
	% (n)	% (n)	% (n)	%
Facility-based LTC, N = 41 166a
Crude	71.5 (29 451)	11.7 (4814)	16.8 (6901)	80.1
Adjustedb, % (95%CI)	71.9 (71.1; 72.6)	11.8 (11.3; 12.2)	16.4 (15.9; 16.9)	75.5 (74.9; 76.1)
Location, crude
Metropolitan	80.3 (22 908)	8.0 (2276)	11.7 (3331)	79.9
Non-metropolitan	53.8 (6482)	16.7 (2011)	29.5 (3546)	80.3
Ownership type, crude
For-profit, crude	77.0 (13 350)	10.3 (1783)	12.7 (2197)	78.9
Not-for-profit	68.5 (15 282)	12.4 (2757)	19.1 (4267)	80.9
Government-run	53.5 (819)	17.9 (274)	28.6 (437)	79.0
Home-based LTC, N = 27 915a
Crude	10.2 (2849)	57.8 (16 145)	32.0 (8921)	77.9
Adjustedb, % (95% CI)	10.1 (9.7; 10.6)	57.9 (57.0; 58.9)	31.5 (30.7; 32.4)	73.6 (72.8; 74.3)
Location, crude
Metropolitan	9.4 (1907)	58.2 (11 768)	32.4 (6546)	78.3
Non-metropolitan	12.2 (942)	56.9 (4377)	30.9 (2373)	76.7
Ownership type, crude
For-profit	9.6 (751)	57.8 (4515)	32.6 (2544)	77.9
Not-for-profit	10.2 (1842)	57.8 (10 479)	32.1 (5815)	78.1
Government-run	13.0 (256)	58.5 (1151)	28.5 (562)	75.5

Abbreviation: CI, confidence interval.

^aAnalyses restricted to older adults who entered care in 2019, stayed in care for at least 180 days, and had at least three GP visits in the 180 days after care entry.

^bEstimates adjusted by age, sex, number of health conditions (comorbidity score), location, and ownership type in a pooled analysis of home-based and facility-based LTC.

In metropolitan areas, 11.7% (3331/28 515) of individuals entering facility-based LTC retained their usual provider compared to 29.5% (3546/12 039) in non-metropolitan areas (Table 2). Of those starting home-based LTC in metropolitan areas, 32.4% (6546/20 221) retained their usual provider compared to 30.9% (2373/7692) in non-metropolitan areas (Table 2). Looking at the results by ownership type in facility-based LTC, retaining one's usual provider was most frequent in government-run LTC (437/1530, 28.6%), followed by not-for-profit (4267/22 306, 19.1%) and for-profit (2197/17 330, 12.7%) LTC (Table 2). In contrast, in home-based LTC, the usual provider retention was highest in people accessing care from for-profit providers (2544/7810, 32.6%), followed by not-for-profit (5815/18 136, 32.1%), and government-run (562/1969, 28.5%) providers (Table 2).

Transitional continuity was similar in the expanded cohort and between those ≤180 and >180 days in care (Supplementary Table S2).

Post-entry continuity

Crude post-entry continuity was comparable between facility- and home-based LTC (UPI = 80.1% in facilities versus 77.9% at home) (Table 2). After adjustment, those in LTC facilities had a 1.9-percentage-point higher probability of seeing their usual provider for any given attendance than those receiving LTC at home (adjusted probabilities 75.5% versus 73.6%, difference = 0.019, 95% CI 0.015–0.022, p < 0.001) (Table 2). Additional adjustment for the number of post-entry primary care attendances using a natural spline produced adjusted usual-provider probabilities of 71.3% (95% CI 70.5%–72.1%) in home-based LTC and 77.4% (95% CI 76.8%–78.1%) in facility-based LTC. Post-entry continuity estimates by location and ownership type were comparable, ranging from 75.5% in government-run home-based LTC to 80.9% in not-for-profit facility-based LTC (Table 2).

Both transitional continuity and post-entry continuity of care estimates were similar to the overall cohort among those with dementia (Supplementary Tables S3–4).

Discussion

Across a national cohort of LTC entrants in 2019, we found a large loss of continuity of primary care during transitions into facility-based LTC (72% changed to a new provider), a smaller but notable loss among those commencing LTC at home (10%), and high post-entry continuity in both settings, with 78%–80% of attendances delivered by a single provider. These results indicate that continuity of care in LTC follows a two-phase pattern: an initial disruption at the point of entry, particularly for facility-based LTC, followed by a high level of continuity after entry. Given the strong associations between continuity of primary care and improved health outcomes, including reduced hospitalizations and mortality [3, 17, 18], this distinction is important. While the transition into LTC facilities remains a principal continuity breakpoint, consistently high post-entry continuity levels indicate that care delivery models are effective at re-establishing relational continuity soon after entry. This suggests that, alongside efforts to preserve continuity across transitions, equal attention should be given to the timely consolidation of care following LTC entry, as this is currently a strength that needs to be valued and maintained across both home-based and facility-based settings.

Nearly three in four facility entrants changed to a new main provider, compared with one in 10 among those starting LTC at home, consistent with prior Australian and international studies [6, 8, 9]. While these disruptions are commonly attributed to residential relocation and limited GP availability in LTC, our findings suggest that these factors alone do not fully explain the observed patterns. If relocation distance were the dominant driver, one would expect lower transitional continuity in non-metropolitan areas, where relocation distances are greater [19]. However, we observed substantially higher rates of switching to new providers in metropolitan areas. This indicates that other factors, including GP availability, differences in primary care delivery models in LTC, or organizational processes within facilities, likely play an important role in shaping transitional continuity. Given the risks associated with continuity disruptions [3], the facility entry period remains a critical window for intervention, with the aim of facilitating transitional continuity where possible, and informational continuity otherwise. Potential levers include standardized handover when the pre-entry GP will not continue (explicit designation of the post-entry provider, structured transfer-of-care information, and a timely first visit [25]); anticipatory planning in home care (encouraging selection of a GP who services the likely future facility); and reducing logistical and financial barriers for community GPs to attend facilities (e.g. telehealth, scheduling support, and remuneration reforms [36]). However, these strategies should be complemented by a clearer focus on managing, rather than entirely preventing, transitional discontinuity, including strengthening informational continuity and ensuring rapid establishment of a new usual provider.

Once established in care, relational continuity was high in both LTC at home and in facilities, with approximately four in five attendances in the 180 days post-entry delivered by a usual provider and little variation by location or ownership type. This finding is notable in light of the substantial disruptions observed during the transition period and suggests that primary care delivery models are effective in consolidating care under a (new) main provider soon after LTC entry. The observed levels of continuity exceed benchmarks reported for community-dwelling older adults in the United States (approximately one in two attendances, although differences in population and healthcare-system structure limit direct comparison [4]) and can be interpreted as the success of the various policies facilitating continuous and person-centred care in the primary care-LTC nexus [26, 37]. Aligning reforms introduced after the study period, such as general practice enrolment [38], may consolidate this success and help sustain high levels of continuity across settings and population subgroups [7].

Limitations

The findings should be interpreted in light of several limitations. First, continuity of care was estimated using national healthcare claims data and practitioner identifiers. Multiple clinicians should not ordinarily be represented by the same identifier, as providers are not permitted to use another healthcare provider's number; however, misclassification may still arise through unauthorized cross-billing or administrative errors. Such errors are considered uncommon and would not be expected to vary systematically across cohorts or strata, making them unlikely to explain the observed relative differences. Second, continuity could not be examined among Aboriginal and Torres Strait Islander individuals (2%) or DVA cardholders (4% in home-based LTC and 11% in facilities), limiting generalizability to these groups. Third, although adjustment and stratification included major individual and facility-level factors expected to influence continuity, residual confounding remains possible because socioeconomic circumstances, frailty, conditions not captured by the RxRisk-V, and characteristics specific to facility-based care, such as facility size, were unavailable or could not be incorporated comparably across care settings. However, estimates were similar in sensitivity analyses stratified by dementia status. Furthermore, data on the primary care delivery models operating within individual facilities were unavailable, precluding assessment of whether transitional and post-entry continuity differed across models. Future research should determine which arrangements most effectively preserve continuity at entry and establish and sustain continuity thereafter.

This study has several strengths. It provides population-level, generalizable estimates of both transitional and post-entry continuity using a large, national cohort. By examining contemporaneous cohorts entering facility-based and home-based LTC, the study enables comparison across care settings with similar underlying health needs. Importantly, by jointly assessing continuity across the transition and the post-entry periods, the study offers a more complete characterization of continuity as a dynamic process rather than a single outcome. Given that continuity of care remains underevaluated in primary healthcare research [39], this study provides novel evidence on how continuity is disrupted and subsequently re-established in LTC settings.

Conclusion

Entry into LTC facilities is associated with substantial disruption of preexisting patient–practitioner relationships, with more than 70% of new residents seeing a new provider after entry compared with approximately 10% among those entering home-based LTC. However, continuity of care is consistently high once individuals are established in care, suggesting that primary care delivery models effectively consolidate care under a usual provider following entry. Importantly, people entering facility-based LTC in metropolitan areas were less likely to keep their provider than those in non-metropolitan areas, suggesting that transitional continuity is not primarily driven by relocation distances and may be amenable to interventions facilitating the integration of primary and LTC. Nevertheless, currently, transitional discontinuity is the norm, and policy efforts should therefore focus not only on preserving continuity where feasible but also on managing transitional discontinuity through structured handover, improved information sharing, and timely establishment of a new usual provider. Monitoring both transitional and post-entry continuity as complementary quality indicators may help ensure that older adults retain the benefits of continuous, trusted care across all LTC settings.

Supplementary material

Supplementary material is available at Family Practice online.

Authors’ contributions

Johannes Schwabe (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Writing—original draft), Maria C. Inacio (Conceptualization, Funding acquisition, Methodology, Resources, Supervision, Validation), Gillian E. Caughey, Maria Crotty, David Roder, Krystal-Lee Nixon, Gillian Harvey, Janet K. Sluggett, Monica Cations, Jyoti Khadka, Megan Corlis, Carolyn Dawkins, Marilyn von Thien (Funding acquisition), Gillian E. Caughey (Resources), Gillian E. Caughey, Maria Crotty, Helena Williams, Andrew Kellie, David Roder, Krystal-Lee Nixon, Gillian Harvey, Janet K. Sluggett, Monica Cations, Tiffany K. Gill, Jyoti Khadka, Megan Corlis, Carolyn Dawkins, Marilyn von Thien, Maria C. Inacio (Supervision). Johannes Schwabe, Brian W. Pulling, Gillian E. Caughey, Maria Crotty, Helena Williams, Andrew Kellie, David Roder, Krystal-Lee Nixon, Gillian Harvey, Janet K. Sluggett, Monica Cations, Tiffany K. Gill, Jyoti Khadka, Megan Corlis, Carolyn Dawkins, Marilyn von Thien, Maria C. Inacio (Writing—review & editing).

Funding

This work is supported by funding from a Medical Research Future Fund Primary Health Care Research Grant (MRFF1200056). J.K.S. is supported by a National Health and Medical Research Council Investigator Grant (GNT2016277). Prof Maria Inacio is supported by the Hospital Research Foundation Mid-Career Fellowship (MCF-27-2019) and a National Health and Medical Research Council Investigator Grant (GNT119378). Prof Gillian Caughey is supported by a National Health and Medical Research Council Investigator Grant (GNT2026400).

Data availability

The data underlying this report are not available for sharing because of restrictions imposed by the ethics and original data custodian approval.

Ethics approvals

This study obtained ethics approvals from the University of South Australia Human Research Ethics Committee (Ref: 200489), the Australian Institute of Health and Welfare Ethics Committee (Ref: EO2022/4/1376), the South Australian Department for Health & Wellbeing Human Research Ethics Committee (Ref: HREC/18/SAH/90) for the inclusion of South Australia, Victorian and Queensland datasets and the New South Wales Population & Health Services Research Ethics Committee (Ref: 2019/ETH12028).