ABSTRACT
Objective
Remote residents have worse health outcomes than metropolitan residents, but whether geography impacts the outcomes of children with prenatal drug exposure (PDE) is uncertain.
Design and Main Outcome Measures
Linked population data was used to compare rates of death, hospitalisation, emergency department (ED) encounters and placement in Out‐of‐Home Care (OOHC) for children with (n = 208 492) and without (n = 1 607 662) smoking, alcohol or substance exposure in NSW, Australia, born between 2001 and 2020. The relationship between remoteness of residence, PDE, service utilisation and death was determined with interaction terms after adjusting for confounders (First Nations status, maternal age, maternal mental illness, social disadvantage). Outcomes are reported as adjusted incidence rate ratios with 95% confidence intervals.
Results
Compared to other NSW children, children with PDE were more likely to die (2.90; 2.33–3.61), be placed in OOHC (52.43; 50.14–54.84), be hospitalised (1.22; 1.10–1.36) and attend ED (1.06; 1.03–1.09). Compared to metropolitan children with PDE, regional children were more likely to die (1.12; 0.78–1.61) but less likely to be hospitalised (0.82; 0.72–0.94), placed in OOHC (0.41; 0.37–0.44) and attend ED (0.93; 0.89–0.98). ED encounters increased up to 2.5‐fold for regional children with PDE from 2001 to 2020.
Conclusions
Children with PDE in regional areas are likely to use fewer hospital resources than their metropolitan counterparts, but are more likely to die, mainly from preventable causes. Equitable access to preventative health and psychosocial support must be prioritised to improve outcomes of these children.
Keywords: health disparate minority, health inequities, healthcare disparities, prenatal exposure delayed effects, substance‐related disorders, vulnerable populations
Highlights
- What's known on this subject
-
○Prenatal drug exposure (PDE) leads to greater mortality and risk of hospital and emergency department admissions in infancy and childhood. People living in regional areas face additional barriers to accessing care and therefore have worse health outcomes than metropolitan residents.
-
○
- What this study adds
-
○Whether geography affects the health and survival of children with prenatal drug exposure (PDE), who are already vulnerable, is uncertain. This study demonstrated that regional children with PDE use fewer resources than their metropolitan counterparts but are more likely to die.
-
○
- How this study might affect research, practice or policy
-
○Despite the Australian Government's funding to close the rural–urban health gap, disparities in maternal drug use and the outcomes of PDE are growing. This study emphasises the need to evaluate models of care to support women and children with PDE in regional and remote areas.
-
○
Abbreviations
- 95% CI
95% confidence interval
- ED
emergency department
- HR
hazard ratio
- ICD‐10‐AM
International Statistical Classification of Disease and Related Problems (10th Edition) Australian Modification
- IRR
incidence rate ratio
- NAS
Neonatal Abstinence Syndrome
- NICU
neonatal intensive care unit
- NSW
New South Wales
- OOHC
out‐of‐home care
- PDE
prenatal drug exposure
1. Introduction
Prenatal drug exposure (PDE) is a critical public health concern [1] firmly associated with adverse pregnancy and neonatal outcomes [2]. Infants with PDE may experience withdrawal, or Neonatal Abstinence Syndrome (NAS) [3], that has the potential to cause serious morbidities and mortality if untreated [4]. While improved recognition and treatment have reduced NAS‐related mortality [5], long‐term outcomes of this rapidly expanding population of children remain concerning [6, 7] due to the direct biological effects of exposure [2], and adverse socioeconomic [5, 8, 9, 10] and environmental [11] circumstances accompanying parental drug use.
We, and others, have shown that PDE and NAS increase the risk of death, hospitalisations and emergency department (ED) admissions in infancy [7, 12, 13] and until childhood [5, 6, 9, 14, 15]. Long‐term support is crucial to reduce adverse outcomes, but this is challenging in marginalised and remote communities with poor access to follow‐up services [5, 9, 16, 17] that are needed for sustained benefit [18].
PDE is increasingly affecting remote, resource‐deprived regions at alarming rates [16], especially in countries like Australia [19] and the United States [20], where the increase in rural rates of NAS was 2–2.5 times higher than in metropolitan areas [16]. Residents in remote areas face disproportionately more barriers to accessing health care services that are paramount to preventing PDE and improving outcomes, including antenatal care [5, 9] and treatment for drug use disorders [16, 17] that necessitate in‐person attendance. Furthermore, people in nonmetropolitan areas are more likely to be financially disadvantaged [21] and have limited health literacy [22, 23]. Given the high proportion of First Nations people living in regional and remote areas, anti‐Indigenous discrimination and poor accessibility to culturally sensitive healthcare also contribute to these disparities [24, 25, 26].
The impact of geographical remoteness on PDE‐related childhood outcomes remains unclear. A retrospective cohort study found that among infants with NAS, rural residence was associated with increased utilisation of outpatient and ED services in the first year of life [7]. Due to difficulties in tracking this large and often marginalised group of children, there remains little information on long‐term outcomes, particularly for those without a NAS diagnosis [27, 28]. No study has evaluated geographical disparities in childhood outcomes after PDE, despite funding initiatives from many governments, including Australia [29], to improve health outcomes in regional and remote areas.
This study used linked population data to assess whether geographical remoteness affected the long‐term health outcomes of infants with PDE, hypothesising that those in regional and remote areas would have reduced access to health services (e.g., hospitalisation, ED presentations) and higher risks of adverse outcomes, including out‐of‐home care (OOHC) placement and death.
2. Methodology
2.1. Study Design and Setting
This population birth cohort study used routinely collected birth, hospitalisation, ED, death and OOHC records of all infants born in NSW, Australia, between 1 July 2001 and 31 December 2020.
2.2. Record Linkage
The Centre for Health Record Linkage (CHeReL) is an independent facility responsible for the linkage of patient records for research and other purposes [30]. Extracts of patient records were linked using probabilistic methods according to names, dates of birth, addresses and hospital identification numbers, which carry a small risk of incorrect linkages (approximately 0.5%) [30]. Each patient was assigned a unique Project‐specific Person Number (PPN), ensuring complete anonymity to researchers. Details of the linked datasets are provided in Table S1.
2.3. Participant Selection and Grouping
As outlined in Figure 1, all births in NSW (n = 1 834 685) were grouped based on a history of PDE after exclusion of stillbirths (n = 11 188), neonatal deaths before discharge (n = 4501) and babies with missing birth discharge data (n = 2842). Infants with PDE (n = 208 492) were grouped according to the type of substance exposure identified in maternal records and the presence of PDE or NAS diagnoses in infant records.
Unexposed Reference Group: No Known History of SUP, PDE or NAS From Infant or Maternal Records (n = 1 607 662)
- A history of substance use in pregnancy (SUP) or PDE (n = 208 492), including
- Maternal smoking only during pregnancy (n = 185 773)
- Maternal alcohol use/dependence only (n = 2771)
- Maternal smoking and alcohol use/dependence (n = 1861)
- Maternal drug use/dependence with/without smoking and/or alcohol use/dependence, but PDE not documented (ICD‐10‐AM F10–19; n = 10 890)
- Child with documented PDE, but no diagnosis of NAS at birth (ICD‐10‐AM P04.4, n = 1254)
- Child diagnosed with NAS at birth (ICD‐10‐AM P96.1, n = 5943)
FIGURE 1.
Participant flow diagram.
Children with PDE were also divided into larger groups, including children exposed to tobacco smoke and/or alcohol (n = 190 405), and children exposed to illicit substances with or without diagnosed PDE or NAS (n = 18 087).
2.4. Definition of Remoteness
Remoteness was categorised according to the Accessibility/Remoteness Index of Australia (ARIA+), which consists of five categories stratifying relative geographic access to service centres, measured by road distance: major city (Category 0), inner regional (Category 1), outer regional (Category 2), remote (Category 3) and very remote areas (Category 4) [31]. Remoteness was also divided into larger categories for some analyses: Metropolitan (ARIA+: 0), Regional (ARIA+: 1 and 2) and Remote (ARIA+: 3 and 4).
2.5. Data Analysis
The effect of remoteness (ARIA+), social disadvantage (Socio‐Economic Indexes for Areas [SEIFA]), First Nations Status, maternal mental illness and maternal age on the incidence of PDE was determined by Poisson regression to calculate the incidence rate ratio (IRR). Poisson regression was also used to assess the effect of PDE on infant demographics, including prematurity, low birth weight, admission to the NICU or special care nursery (SCN), as well as outcomes beyond discharge such as death (at any age, ≤ 28 days, 29–365 days, 1–5 years, 6–10 years, 11–20 years) and placement into OOHC (at any age, ≤ 6 months and ≤ 12 months). Linear regression was used to compare birth admission length, and Cox regression was used to assess time to death, with hazard ratios (HRs) calculated. Because a large proportion of participants had no further admissions beyond birth, a zero‐inflated Poisson regression was used to compare the number of hospital and ED encounters across age groups. Negative binomial regression was used for total days in hospital and total cost of hospitalisations due to overdispersion.
Choropleth maps of NSW, divided into Statistical Areas Level 2 (SA2s) [31] were constructed to assess temporal and geographical trends within each PDE group. Within each SA2, Poisson regression was used to estimate the crude annual change in the incidence of PDE and in cohort outcomes, including death, OOHC placements, hospitalisations and ED encounters. The resultant IRRs were then merged with the dataset containing polygon coordinates for the SA2s, and maps were created using the SPMAP [32] feature in Stata 18.0. Each area was coloured according to the IRR values, with blue indicating a decrease in the rate over the study period, red indicating an increase and white indicating insufficient data for regression.
The interaction between remoteness and PDE in influencing clinical outcomes was evaluated using Poisson regression. Interactions were synergistic (IRR > 1) when the combined effect of remoteness and PDE surpassed the sum of each independently, and antagonistic (IRR < 1) when the combined effect was less than the sum.
Adjustment for confounders was based on the directed acyclic graph depicted in Figure S1, including First Nations Status, remoteness (ARIA+), social disadvantage (SEIFA), young maternal age (under 20 years), serious mental illness in the mother and calendar year of birth.
Data were presented primarily as percentages (%), adjusted IRR (aIRR) with 95% confidence interval (CI) and median (interquartile range [IQR]) for skewed data or mean (standard deviation [SD]) for normally distributed data.
All statistical analyses were performed using Stata (Basic edition; Version 18.0, College Station, TX: StataCorp LLC).
3. Ethics Statement
Authorised by NSW Human Research Ethics Committee on 29/06/2020 (2019/ETH12716) and the Aboriginal Health and Medical Research Council (AHMRC; 19 July 2022), 2019/ETH 12716. Personnel amendment authorised on 24/02/2023. Manuscripts were approved by the Department of Communities and Justice NSW and the AHMRC before submission.
4. Results
4.1. Maternal and Infant Demographics
Mothers with SUP were more likely to be younger, of First Nations background, have a mental health disorder, or belong to the lowest SEIFA quintiles (Table 1). Of all children exposed to substances, only 2.85% received a NAS diagnosis, with rates higher in metropolitan (3.56%) than remote areas (0.40%). Infants with any substance exposure were more likely to be premature and have low birth weight (Table 2). Those with documented PDE or NAS were more likely to be admitted to a SCN or NICU (Group 5: 4.63; 95% CI 4.30–4.99; Group 6: 4.73; 95% CI 4.57–4.89) and required longer birth admissions.
TABLE 1.
Maternal characteristics, by type of SUP and documentation of PDE or NAS.
| Characteristic | Unexposed group a (n = 1 607 662) | Group 1: Maternal smoking during pregnancy (n = 185 773) | Group 2: Maternal alcohol use/dependence (n = 2771) | Group 3: Maternal smoking during pregnancy and maternal alcohol use/dependence (n = 1861) | Group 4: Maternal drug use/dependence±smoking and alcohol use/dependence, but no documented PDE (n = 10 890) | Group 5: Infant with PDE (ICD‐10‐AM P04.4) but not NAS (n = 1254) | Group 6: Infant with NAS (ICD‐10‐AM P96.1) (n = 5943) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n (%) | n (%) | aIRR c , d (95% CI) | n (%) | aIRR c , d (95% CI) | n (%) | aIRR c , d (95% CI) | n (%) | aIRR c , d (95% CI) | n (%) | aIRR c , d (95% CI) | n (%) | aIRR c , d (95% CI) | ||
| Young mother (< 20 years at delivery) | 33 789 (2.10) | 16 696 (8.99) | 2.11 (2.07–2.14) | 83 (3.00) | 1.01 (0.81–1.26) | 131 (7.04) | 1.31 (1.09–1.57 = 8) | 1216 (11.17) | 2.46 (2.31–2.62) | 92 (7.34) | 1.44 (1.16–1.80) | 256 (4.31) | 0.83 (0.73–0.95) | |
| History of serious mental illness b | 19 308 (1.18) | 3612 (1.94) | 1.30 (1.26–1.35) | 651 (23.49) | 21.05 (19.27–23.00) | 265 (14.24) | 10.64 (9.33–12.13) | 1494 (13.72) | 10.04 (9.50–10.61) | 69 (5.50) | 3.64 (2.85–4.64) | 387 (6.51) | 4.40 (3.97–4.89) | |
| Aboriginal and/or Torres Strait Islander heritage | 33 266 (2.07) | 25 727 (13.85) | 4.45 (4.40–4.51) | 147 (5.30) | 1.55 (1.31–1.83) | 399 (21.44) | 7.56 (6.77–8.45) | 2164 (19.87) | 6.87 (6.56–7.20) | 303 (24.16) | 8.83 (7.76–10.05) | 1249 (21.02) | 7.37 (6.93–7.85) | |
| SEIFA quintile e | 1 | 262 355 (16.51) | 35 678 (19.43) | 2.24 (2.20–2.29) | 287 (10.42) | 0.47 (0.41–0.54) | 348 (18.93) | 2.16 (1.76–2.65) | 1949 (18.04) | 2.34 (2.15–2.54) | 254 (20.45) | 2.98 (2.35–3.78) | 1460 (24.83) | 4.20 (3.75–4.70) |
| 2 | 473 146 (29.78) | 75 607 (41.18) | 2.45 (2.41–2.50) | 834 (30.28) | 0.74 (0.67–0.83) | 778 (42.33) | 2.62 (2.17–3.16) | 4442 (41.13) | 2.76 (2.55–2.97) | 443 (35.67) | 2.84 (2.27–3.56) | 2177 (37.03) | 3.33 (2.99–3.72) | |
| 3 | 269 955 (16.99) | 35 836 (19.52) | 2.12 (2.08–2.16) | 451 (16.38) | 0.71 (0.63–0.81) | 344 (18.72) | 2.09 (1.70–2.56) | 2202 (20.39) | 2.43 (2.24–2.63) | 279 (22.46) | 3.14 (2.48–3.98) | 1178 (20.04) | 3.14 (2.80–3.53) | |
| 4 | 273 556 (17.22) | 22 234 (12.11) | 1.61 (1.58–1.65) | 555 (20.15) | 0.95 (0.85–1.06) | 235 (12.79) | 1.80 (1.46–2.23) | 1390 (12.87) | 1.79 (1.64–1.95) | 172 (13.85) | 2.00 (1.56–2.57) | 676 (11.50) | 1.88 (1.66–2.13) | |
| 5 | 309 990 (19.51) | 14 257 (7.76) | — | 627 (22.77) | — | 133 (7.24) | — | 818 (7.57) | — | 94 (7.57) | — | 388 (6.60) | — | |
| ARIA+ category f | 0 | 1 273 067 (80.60) | 108 601 (59.36) | — | 2094 (76.28) | — | 1035 (56.31) | — | 6957 (64.62) | — | 952 (76.84) | — | 4411 (75.50) | — |
| 1 | 230 052 (14.57) | 51 249 (28.01) | 2.03 (2.00–2.05) | 510 (18.58) | 1.27 (1.15–1.40) | 567 (30.85) | 2.35 (2.12–2.61) | 2887 (26.82) | 1.75 (1.67–1.83) | 231 (18.64) | 1.02 (0.88–1.18) | 1200 (20.54) | 1.15 (1.07–1.22) | |
| 2 | 69 736 (4.42) | 20 328 (11.11) | 2.36 (2.32–2.40) | 124 (4.52) | 1.06 (0.88–1.28) | 180 (9.79) | 2.20 (1.87–2.59) | 824 (7.65) | 1.48 (1.38–1.60) | 45 (3.63) | 0.59 (0.44–0.80) | 219 (3.75) | 0.58 (0.51–0.67) | |
| 3 | 5295 (0.34) | 1957 (1.07) | 2.78 (2.66–2.91) | 7 (0.26) | 0.82 (0.39–1.73) | 24 (1.31) | 3.65 (2.43–5.48) | 64 (0.59) | 1.43 (1.12–1.83) | 5 (0.40) | 0.80 (0.33–1.93) | 4 (0.07) | 0.12 (0.05–0.33) | |
| 4 | 1296 (0.08) | 816 (0.45) | 4.04 (3.77–4.33) | 10 (0.36) | 4.89 (2.60–9.17) | 32 (1.74) | 17.73 (12.32–25.52) | 34 (0.32) | 2.75 (1.96–3.86) | 6 (0.48) | 3.17 (1.41–7.13) | 8 (0.14) | 0.74 (0.37–1.49) | |
Note: Data expressed as n (%), unless otherwise stated.
Abbreviations: 95% CI, 95% confidence interval; aIRR, adjusted incidence rate ratio; ARIA+, Accessibility/Remoteness Index of Australia (measure of remoteness based on road distances to service centres); ICD‐10‐AM, International Classification of Diseases, Tenth Revision, Australian Modification; NAS, Neonatal Abstinence Syndrome; PDE, prenatal drug exposure; SEIFA, Socio‐Economic Indexes for Areas (a measure of social disadvantage in a small contained geographical area); SUP, substance use in pregnancy.
Children with no maternal smoking during pregnancy, no maternal alcohol use/dependence, no maternal drug use/dependence, no PDE and no NAS.
Recorded in the last hospital admission or episode of mental health care in an ambulatory care setting prior to birth.
Risk of PDE compared to mothers over 20 years of age, those without a history of serious mental illness, those who are.
Adjustment for young mother: Aboriginal and/or Torres Strait Islander mother, SEIFA, ARIA+; adjustment for serious mental illness in the mother: aboriginal and/or Torres Strait Islander mother, SEIFA, ARIA+, young mother (aged < 20 years); adjustment for SEIFA: ARIA+; adjustment for ARIA+: SEIFA.
5th SEIFA quintile was defined as the base level for regression.
ARIA+ categories: Major Cities (0), Inner Regional (1), Outer Regional (2), Remote (3), Very Remote (4); ARIA+ Category 0 was defined as the base level for regression.
TABLE 2.
Infant characteristics and outcomes, by type of SUP and documentation of PDE or NAS.
| Level of care required at birth | Unexposed group a (n = 1 607 662) | Group 1: Maternal smoking during pregnancy (n = 185 773) | Group 2: Maternal alcohol use/dependence (n = 2771) | Group 3: Maternal smoking during pregnancy and maternal alcohol use/dependence (n = 1861) | Group 4: Maternal drug use/dependence±smoking and alcohol use/dependence, but no documented PDE (n = 10 890) | Group 5: Infant with PDE (ICD‐10‐AM P04.4) but not NAS (n = 1254) | Group 6: Infant with NAS (ICD‐10‐AM P96.1) (n = 5943) | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n (%) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | |
| Prematurity (< 37 weeks' gestation) | 101.49 (6.29) | 17 305 (9.32) | 1.44 (1.42–1.47) | 229 (8.26) | 1.26 (1.11–1.44) | 258 (13.86) | 2.07 (1.83–2.34) | 1820 (16.71) | 2.49 (2.37–2.61) | 325 (25.92) | 3.80 (3.41–4.24) | 1284 (21.61) | 3.18 (3.01–3.37) |
| Low birth weight (< 2500 g) | 80 986 (5.04) | 18 770 (10.10) | 1.97 (1.93–2.00) | 207 (7.47) | 1.47 (1.28–1.68) | 303 (16.28) | 3.12 (2.79–3.50) | 1952 (17.92) | 3.35 (3.20–3.51) | 424 (33.81) | 6.19 (5.63–6.82) | 1427 (24.01) | 4.43 (4.20–4.67) |
| Admission to a NICU or SCN | 172 850 (14.69) | 27 676 (18.50) | 1.21 (1.19–1.22) | 391 (16.66) | 1.07 (0.97–1.18) | 389 (24.31) | 1.50 (1.35–1.66) | 2650 (29.58) | 1.85 (1.78–1.92) | 728 (73.39) | 4.63 (4.30–4.99) | 3589 (75.04) | 4.73 (4.57–4.89) |
| Length of birth admission | Median (IQR) | Median (IQR) | aβ b , c (95% CI) | Median (IQR) | aβ b , c (95% CI) | Median (IQR) | aβ b , c (95% CI) | Median (IQR) | aβ b , c (95% CI) | Median (IQR) | aβ b , c (95% CI) | Median (IQR) | aβ b , c (95% CI) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Median length of stay | 3 (2–4) | 2 (1–4) | −0.28 (−0.31 to −0.25) | 3 (2–5) | 0.32 (0.09–0.56) | 3 (2–4) | 0.50 (0.21–0.79) | 3 (2–5) | 1.21 (1.10–1.33) | 6 (4–10) | 5.22 (4.87–5.57) | 9 (6–17) | 9.23 (9.07–9.39) |
| Death | n (%) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Any age | 3009 (0.19) | 864 (0.47) | 1.94 (1.79–2.10) | 12 (0.43) | 1.87 (1.06–3.30) | 21 (1.13) | 4.03 (2.62–6.21) | 68 (0.62) | 2.37 (1.85–3.03) | 12 (0.96) | 4.04 (2.29–7.14) | 60 (1.01) | 4.12 (3.22–5.41) |
| < 28 days | 678 (0.04) | 171 (0.09) | 1.85 (1.55–2.22) | 1 (0.04) | 0.79 (0.11–5.66) | 4 (0.21) | 3.88 (1.44–10.45) | 15 (0.14) | 2.66 (1.58–4.48) | 1 (0.08) | 1.58 (0.22–10.23) | 5 (0.08) | 1.68 (0.70–4.07) |
| 1 month to 1 year | 1006 (0.06) | 393 (0.21) | 2.75 (2.42–3.11) | 1 (0.04) | 0.49 (0.07–3.49) | 14 (0.75) | 8.64 (5.06–14.74) | 26 (0.24) | 2.71 (1.81–4.06) | 7 (0.56) | 7.08 (3.36–14.92) | 35 (0.59) | 7.28 (5.15–10.30) |
| 1–5 years | 775 (0.05) | 199 (0.11) | 1.72 (1.46–2.03) | 5 (0.18) | 2.96 (1.22–7.17) | 2 (0.11) | 1.47 (0.37–5.92) | 16 (0.15) | 2.20 (1.33–3.64) | 1 (0.08) | 1.32 (0.19–9.39) | 10 (0.17) | 2.77 (1.48–5.19) |
| 6–10 years | 267 (0.02) | 43 (0.02) | 0.97 (0.69–1.36) | 3 (0.11) | 5.36 (1.70–16.91) | 1 (0.05) | 2.06 (0.29–14.81) | 5 (0.05) | 1.93 (0.79–4.73) | 2 (0.16) | 7.62 (1.88–30.82) | 5 (0.08) | 4.00 (1.64–9.77) |
| 11–20 years | 283 (0.02) | 58 (0.03) | 1.11 (0.82–1.48) | 2 (0.07) | 2.17 (0.53–8.85) | 0 (0.00) | – e | 6 (0.06) | 1.69 (0.74–3.85) | 1 (0.08) | 2.99 (0.42–21.39) | 5 (0.08) | 2.99 (1.23–7.31) |
| Survival analysis | – | aHR (95% CI) | aHR (95% CI) | aHR (95% CI) | aHR (95% CI) | aHR (95% CI) | aHR (95% CI) |
|---|---|---|---|---|---|---|---|
| Time to death | — | 1.95 (1.79–2.11) | 1.87 (1.06–3.30) | 4.08 (2.65–6.28) | 2.38 (1.86–3.04) | 4.06 (2.30–7.17) | 4.19 (3.23–5.43) |
| OOHC | n (%) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) | n (%) | aIRR b , c (95% CI) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Any age | 7056 (0.44) | 11 493 (6.19) | 10.13 (9.82–10.46) | 146 (5.27) | 9.25 (7.84–10.92) | 405 (21.76) | 29.28 (26.44–32.43) | 2480 (22.77) | 30.19 (28.76–31.70) | 487 (38.84) | 55.73 (50.76–61.18) | 2369 (39.86) | 59.15 (56.34–62.11) |
| < 6 months | 1524 (0.09) | 2351 (1.27) | 11.44 (10.68–12.25) | 24 (0.87) | 7.87 (5.25–11.80) | 91 (4.89) | 39.11 (31.48–48.58) | 566 (5.20) | 36.97 (33.35–40.98) | 244 (19.46) | 139.77 (121.67–160.56) | 1206 (20.29) | 153.96 (141.98–166.94) |
| < 12 months | 1994 (0.12) | 3146 (1.69) | 11.47 (10.80–12.17) | 33 (1.19) | 8.03 (5.65–11.41) | 122 (6.56) | 39.48 (32.74–47.60) | 737 (6.77) | 36.26 (33.13–39.69) | 272 (21.69) | 118.16 (103.77–134.54) | 1359 (22.87) | 132.05 (122.67–142.14) |
| Measure of healthcare utilisation | Median (IQR) | Median (IQR) | aIRR b , c (95% CI) | Median (IQR) | aIRR b , c (95% CI) | Median (IQR) | aIRR b , c (95% CI) | Median (IQR) | aIRR b , c (95% CI) | Median (IQR) | aIRR b , c (95% CI) | Median (IQR) | aIRR b , c (95% CI) | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Hospital Ad. | 0–1 year | 0 (0–0) | 0 (0–1) | 1.19 (1.17–1.21) | 0 (0–0) | 0.96 (0.88–1.06) | 0 (0–1) | 1.33 (1.19–1.49) | 0 (0–1) | 1.12 (1.06–1.19) | 0 (0–1) | 1.44 (1.28–1.63) | 0 (0–1) | 1.34 (1.26–1.42) |
| 1–5 years | 0 (0–1) | 0 (0–1) | 1.06 (1.04–1.09) | 0 (0–1) | 0.94 (0.86–1.03) | 0 (0–1) | 1.00 (0.89–0.12) | 0 (0–1) | 1.12 (0.98–1.27) | 0 (0–1) | 1.10 (0.94–1.28) | 0 (0–1) | 1.22 (1.09–1.36) | |
| 6–10 years | 0 (0–0) | 0 (0–0) | 1.11 (1.05–1.18) | 0 (0–0) | 0.98 (0.77–1.24) | 0 (0–0) | 1.20 (0.81–1.77) | 0 (0–0) | 1.24 (0.96–1.60) | 0 (0–0) | 1.23 (0.68–2.23) | 0 (0–1) | 1.37 (1.03–1.84) | |
| 11–15 years | 0 (0–0) | 0 (0–0) | 1.08 (1.01–1.15) | 0 (0–0) | 1.03 (0.83–1.26) | 0 (0–0) | 0.70 (0.58–0.85) | 0 (0–0) | 1.22 (0.99–1.49) | 0 (0–1) | 0.88 (0.70–1.12) | 0 (0–1) | 1.58 (1.21–2.06) | |
| ED Ad. | 0–1 year | 0 (0–1) | 0 (0–1) | 1.23 (1.22–1.24) | 0 (0–1) | 1.10 (1.01–1.20) | 0 (0–1) | 1.30 (1.18–1.42) | 0 (0–1) | 1.15 (1.11–1.20) | 0 (0–1) | 1.04 (0.94–1.16) | 0 (0–1) | 1.22 (1.16–1.29) |
| 1–5 years | 1 (0–3) | 1 (0–3) | 1.13 (1.12–1.14) | 1 (0–3) | 1.02 (0.96–1.08) | 1 (0–3) | 1.10 (1.03–1.18) | 1 (0–3) | 1.04 (1.01–1.07) | 1 (0–3) | 1.02 (0.93–1.12) | 1 (0–3) | 1.05 (1.01–1.10) | |
| 6–10 years | 0 (0–1) | 1 (0–2) | 1.15 (1.14–1.17) | 1 (0–2) | 1.10 (1.02–1.19) | 1 (0–2) | 1.20 (1.10–1.32) | 1 (0–2) | 1.15 (1.10–1.20) | 1 (0–2) | 0.94 (0.83–1.06) | 0 (0–2) | 1.06 (0.99–1.15) | |
| 11–15 years | 0 (0–2) | 1 (0–3) | 1.31 (1.29–1.34) | 1 (0–2) | 1.23 (1.10–1.37) | 1 (0–2) | 1.22 (1.06–1.41) | 1 (0–3) | 1.44 (1.33–1.57) | 1 (0–3) | 1.75 (1.21–2.55) | 1 (0–3) | 1.93 (1.64–2.29) | |
| LOS | 0–1 year | 4 (2–5) | 3 (2–5) | 1.03 (1.03–1.04) | 4 (2–5) | 1.07 (1.03–1.10) | 3 (2–6) | 1.28 (1.23–1.34) | 4 (2–7) | 1.47 (1.44–1.49) | 7 (5–14) | 2.68 (2.55–2.81) | 11 (7–22) | 3.36 (3.28–3.43) |
| 1–5 years | 0 (0–0) | 0 (0–1) | 1.15 (1.13–1.17) | 0 (0–1) | 0.95 (0.84–1.07) | 0 (0–1) | 1.10 (0.95–1.27) | 0 (0–1) | 1.25 (1.18–1.34) | 0 (0–1) | 1.50 (1.24–1.82) | 0 (0–1) | 1.41 (1.29–1.53) | |
| 6–10 years | 0 (0–0) | 0 (0–0) | 1.16 (1.13–1.19) | 0 (0–0) | 1.05 (0.86–1.30) | 0 (0–0) | 1.12 (0.88–1.43) | 0 (0–0) | 1.29 (1.16–1.44) | 0 (0–0) | 1.36 (0.97–1.90) | 0 (0–0) | 1.44 (1.24–1.67) | |
| 11–15 years | 0 (0–0) | 0 (0–0) | 1.25 (1.19–1.32) | 0 (0–0) | 1.80 (1.26–2.58) | 0 (0–0) | 0.79 (0.51–1.21) | 0 (0–0) | 2.04 (1.69–2.75) | 0 (0–0) | 1.53 (0.86–2.75) | 0 (0–0) | 2.03 (1.56–2.64) | |
| Cost d | 0–1 year | 3570.75 (3570.75–6171.39) | 3570.75 (3570.75–6488.72) | 1.22 (1.21–1.23) | 3570.75 (3570.75–6171.39) | 1.13 (1.08–1.18) | 3570.75 (3570.75–7675.20) | 1.38 (1.30–1.46) | 3570.75 (3570.75–8435.05) | 1.49 (1.46–1.53) | 6171.39 (3570.75–13617.08) | 2.35 (2.20–2.52) | 11598.36 (9335.07–18408.68) | 2.65 (2.56–2.73) |
| 1–5 years | 0 (0–2595.32) | 0 (0–3028.12) | 1.12 (1.09–1.15) | 0 (0–3028.12) | 0.97 (0.79–1.20) | 0 (0–3028.12) | 1.08 (0.84–1.39) | 0 (0–3028.12) | 1.21 (1.09–1.35) | 0 (0–3608.68) | 1.41 (1.02–1.96) | 0 (0–3520.08) | 1.34 (1.15–1.55) | |
| 6–10 years | 0 (0–0) | 0 (0–0) | 1.14 (1.09–1.20) | 0 (0–0) | 1.01 (0.71–1.43) | 0 (0–0) | 1.21 (0.80–1.83) | 0 (0–0) | 1.32 (1.10–1.58) | 0 (0–0) | 1.36 (0.77–2.41) | 0 (0–0) | 1.49 (1.15–1.92) | |
| 11–15 years | 0 (0–0) | 0 (0–0) | 1.27 (1.17–1.39) | 0 (0–0) | 1.39 (0.76–2.53) | 0 (0–0) | 0.99 (0.49–1.99) | 0 (0–0) | 1.62 (1.18–2.21) | 0 (0–2282.40) | 1.64 (0.62–4.32) | 0 (0–0) | 2.06 (1.33–3.18) | |
Note: Data expressed as n (%), unless otherwise stated.
Abbreviations: 95% CI, 95% confidence interval; Ad., admissions; aIRR, adjusted incidence rate ratio; aβ, adjusted regression coefficient.; ED, emergency department; HR, hazard ratio; ICD‐10‐AM, International Classification of Diseases, Tenth Revision, Australian Modification; IQR, interquartile range; LOS, length of stay; NAS, Neonatal Abstinence Syndrome; NICU, neonatal intensive care unit; OOHC, out‐of‐home care; PDE, prenatal drug exposure; SCN, special care nursery; SUP, substance use in pregnancy.
Children with no maternal smoking during pregnancy, no maternal alcohol use/dependence, no maternal drug use/dependence, no PDE and no NAS.
Risk compared to Unexposed Group (children with no maternal smoking during pregnancy, no maternal alcohol use/dependence, no maternal drug use/dependence, no PDE and no NAS).
Adjusted for: aboriginal and/or Torres Strait Islander mother; serious mental illness in the mother; Accessibility/Remoteness Index of Australia (ARIA+) (measure of remoteness based on road distances to service centres); Socio‐Economic Indexes for Areas (SEIFA) (a measure of social disadvantage in a small contained geographical area); young mother (aged < 20 years); calendar year.
Inpatient hospital cost for each admission was determined according to the Australian Refined Diagnosis Related Group (AR‐DRG) cost.
Not calculable, very low death rate.
4.2. Geographical Distribution of PDE
Maternal smoking and/or alcohol use significantly increased with level of remoteness (Group 3: aIRR 17.73; 95% CI 12.32–25.52). Illicit substance exposure without documented PDE increased in all regional and remote areas, peaking in very remote areas (Group 4: aIRR 2.75; 95% CI 1.96–3.86). Documented PDE also peaked in very remote NSW; however, it decreased in outer regional and remote areas (Group 5 ARIA+2: aIRR 0.59; 95% CI 0.44–0.80). NAS increased slightly in inner regional NSW and decreased with remoteness, most notably in remote areas (Group 6 ARIA+3: aIRR 0.12; 95% CI 0.05–0.33) (Table 1).
4.3. Temporal Change in Geographical Distribution of PDE
The temporal change in maternal substance use, PDE and NAS from 2001 to 2020 is presented in Figure 2A, with NSW on the left and Sydney on the right. Maternal smoking and alcohol use generally decreased across most areas of NSW, except for some outer regional areas where maternal alcohol use increased. Maternal exposure to illicit substances without diagnosed PDE or NAS decreased in metropolitan areas but increased in regional and remote NSW. Rates of diagnosed PDE and NAS notably increased in non‐metropolitan areas.
FIGURE 2.
Choropleth maps demonstrating the changes in the incidence of PDE and rates of adverse childhood outcomes across NSW. (A) Incidence, (B) death, (C), OOHC provision, (D) hospital admissions and (E) ED admissions.
4.4. Impact of PDE on Childhood Outcomes and Temporal Changes
Cohort outcomes, including mortality, placement into OOHC, hospitalisations and ED encounters, are detailed in Table 2. The temporal changes in the incidence of these outcomes across NSW are presented in Figure 2.
4.4.1. Mortality
Of the total 4046 deaths post‐discharge, 1037 (25.6%) occurred in children with PDE. Of all deaths, 45.3% occurred outside of metropolitan areas for PDE children, compared with 23.2% for children without PDE. This finding was overall not statistically significant, likely due to the relatively small number of deaths. Most deaths occurred in the first year of life. Children with a diagnosis of NAS (Group 6) were at the highest overall risk of death, while children diagnosed with PDE but not NAS (Group 5) were at the highest risk between 6 and 10 years of age (Table 2). Risk of death increased between 2001 and 2020 in children of mothers who smoked in certain regions (e.g., remote Far West, Figure 2B).
4.4.2. Placement in OOHC
PDE was strongly associated with OOHC placement, particularly for infants with NAS (Group 6, Table 2). OOHC placements increased overall for regional centres between 2001 and 2020, particularly in remote and very remote areas, whereas they decreased in most metropolitan areas (Figure 2C).
4.4.3. Hospitalisations and ED Encounters
PDE was primarily associated with increased risk of hospital and ED encounters throughout childhood up to 15 years of age (Table 2), especially in children with NAS between 11 and 15 years. Children with PDE also spent a greater number of days in hospital and had higher total costs of hospital admissions.
Overall risk of hospitalisations and ED encounters, however, decreased over time for children without PDE. In remote areas, hospitalisations decreased, and ED encounters increased 2.5‐fold for those with PDE (Figure 2D,E).
4.5. Interaction Between Geographical Remoteness and PDE
PDE and remoteness independently increased the risk of mortality, OOHC, hospitalisation and ED encounters with varying interaction effects. In their first year of life, remote infants with PDE had more hospitalisations and ED encounters compared to metropolitan infants with PDE. Between 1 and 5 years, regional children with PDE were 12% more likely to die than those living in major cities but had lower rates of hospitalisation, ED presentations and placements in OOHC. Remote children with PDE used even fewer resources but were 9% more likely to die compared to metropolitan counterparts. Children diagnosed with NAS in the neonatal period were 68% more likely to die if residing outside major cities. Effects were consistent from 1 to 15 years of age (Table 3).
TABLE 3.
Interaction between geographical remoteness and PDE in influencing cohort outcomes.
| Outcome | Effect of PDE | Effect of geographical remoteness | Interaction terms | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Maternal smoking during pregnancy and/or alcohol use/dependence a (n = 190 405) | Maternal drug use/dependence and/or documented PDE and/or diagnosed NAS b (n = 18 087) | Regional (ARIA+ Categories 1 and 2) (n = 378 152) | Remote (ARIA+ Categories 3 and 4) (n = 9558) | Regional * Exposure to tobacco smoke and/or alcohol (n = 72 958) | Remote * Exposure to tobacco smoke and/or alcohol (n = 2846) | Regional * Exposure to drugs and/or documented PDE and/or diagnosed NAS (n = 5406) | Remote * Exposure to drugs and/or documented PDE and/or diagnosed NAS (n = 121) | ||||||||||
| aIRR c , d (95% CI) | p | aIRR c , d (95% CI) | p | aIRR c , d (95% CI) | p | aIRR c , d (95% CI) | p | aIRR c , d (95% CI) | p | aIRR c , d (95% CI) | p | aIRR c , d (95% CI) | p | aIRR c , d (95% CI) | p | ||
| Death | 1.98 (1.79–2.19) | < 0.001 | 2.90 (2.33–3.61) | < 0.001 | 1.13 (1.04–1.24) | 0.005 | 1.38 (0.91–2.11) | 0.133 | 0.99 (0.85–1.17) | 0.943 | 0.84 (0.46–1.54) | 0.569 | 1.12 (0.78–1.61) | 0.529 | 1.09 (0.25–4.72) | 0.908 | |
| OOHC | 12.60 (12.14–13.09) | < 0.001 | 52.43 (50.14–54.84) | < 0.001 | 1.59 (1.51–1.67) | < 0.001 | 1.44 (1.12–1.84) | 0.004 | 0.51 (0.48–0.54) | < 0.001 | 0.36 (0.27–0.47) | < 0.001 | 0.41 (0.37–0.44) | < 0.001 | 0.38 (0.27–0.55) | < 0.001 | |
| Hospital Ad. | 0–1 years | 1.21 (1.18–1.23) | < 0.001 | 1.23 (1.17–1.30) | < 0.001 | 1.06 (1.04–1.08) | < 0.001 | 1.09 (1.01–1.17) | 0.019 | 0.95 (0.93–0.98) | 0.001 | 1.05 (0.94–1.17) | 0.423 | 0.99 (0.92–1.07) | 0.760 | 1.11 (0.70–1.77) | 0.660 |
| 1–5 years | 1.06 (1.03–1.09) | < 0.001 | 1.22 (1.10–1.36) | < 0.001 | 0.95 (0.93–0.97) | < 0.001 | 1.03 (0.93–1.16) | 0.556 | 1.00 (0.96–1.04) | 0.943 | 0.99 (0.86–1.14) | 0.877 | 0.82 (0.72–0.94) | 0.005 | 0.70 (0.52–0.94) | 0.016 | |
| 6–10 years | 1.17 (1.09–1.27) | < 0.001 | 1.32 (1.02–1.69) | 0.032 | 0.91 (0.87–0.95) | < 0.001 | 0.91 (0.78–1.06) | 0.217 | 0.85 (0.78–0.93) | 0.001 | 0.81 (0.67–0.99) | 0.042 | 0.92 (0.64–1.23) | 0.560 | 0.77 (0.47–1.26) | 0.294 | |
| 11–15 years | 1.10 (1.01–1.19) | 0.023 | 1.27 (1.05–1.55) | 0.015 | 0.92 (0.86–0.98) | 0.012 | 0.92 (0.71–1.19) | 0.535 | 0.95 (0.85–1.06) | 0.337 | 0.85 (0.62–1.15) | 0.288 | 1.15 (0.82–1.61) | 0.426 | 1.00 (0.62–1.59) | 0.984 | |
| ED Ad. | 0–1 years | 1.24 (1.23–1.26) | < 0.001 | 1.18 (1.13–1.23) | < 0.001 | 1.26 (1.25–1.27) | < 0.001 | 0.96 (0.90–1.02) | 0.186 | 0.96 (0.95–0.98) | < 0.001 | 0.88 (0.79–0.99) | 0.031 | 0.95 (0.90–1.01) | 0.108 | 1.08 (0.73–1.59) | 0.708 |
| 1–5 years | 1.12 (1.11–1.13) | < 0.001 | 1.06 (1.03–1.09) | < 0.001 | 1.33 (1.32–1.33) | < 0.001 | 1.06 (1.01–1.11) | 0.026 | 1.00 (0.98–1.01) | 0.778 | 0.86 (0.79–0.94) | 0.001 | 0.93 (0.89–0.98) | 0.003 | 0.66 (0.41–1.07) | 0.091 | |
| 6–10 years | 1.17 (1.15–1.19) | < 0.001 | 1.13 (1.07–1.18) | < 0.001 | 1.39 (1.37–1.40) | < 0.001 | 1.25 (1.17–1.34) | < 0.001 | 0.95 (0.93–0.97) | < 0.001 | 0.96 (0.86–1.07) | 0.429 | 0.95 (0.88–1.02) | 0.162 | 0.69 (0.45–1.07) | 0.095 | |
| 11–15 years | 1.36 (1.33–1.40) | < 0.001 | 1.68 (1.52–1.85) | < 0.001 | 1.39 (1.37–1.42) | < 0.001 | 1.46 (1.36–1.57) | < 0.001 | 0.91 (0.87–0.94) | < 0.001 | 0.87 (0.77–0.98) | 0.025 | 0.87 (0.73–1.04) | 0.135 | 0.94 (0.62–1.43) | 0.776 | |
Abbreviations: 95% CI, 95% confidence interval; Ad., admissions; aIRR, adjusted incidence rate ratio; ARIA+, Accessibility/Remoteness Index of Australia (measure of remoteness based on road distances to service centres); ED, emergency department; HR, hazard ratio; IQR, interquartile range; LOS, length of stay; NAS, Neonatal Abstinence Syndrome; OOHC, out‐of‐home care; PDE, prenatal drug exposure.
Recorded in the last hospital admission or episode of mental health care in an ambulatory care setting prior to birth.
ICD‐10‐AM (International Classification of Diseases, Tenth Revision, Australian Modification) codes P04.4 for PDE and P96.1 for NAS.
Risk compared to Unexposed Group (children with no maternal smoking during pregnancy, no maternal alcohol use/dependence, no maternal drug use/dependence, no PDE and no NAS).
Adjusted for: aboriginal and/or Torres Strait Islander mother; serious mental illness in the mother; Socio‐Economic Indexes for Areas (SEIFA) (a measure of social disadvantage in a small contained geographical area); young mother (aged < 20 years); calendar year.
5. Discussion
This population‐based cohort study is the first report of geographical disparities in the long‐term outcomes of PDE at a population level. Results confirm the association between PDE and adverse health outcomes [5, 9] and highlight that a growing proportion of children with PDE are from regional and remote areas, where they access fewer resources and are more likely to die.
In keeping with existing literature, children from remote areas are more likely to be affected by PDE [16, 20, 33], where they face poor access to healthcare services [5, 9, 16, 17], lower health literacy [21, 22, 23] and cultural barriers [24, 25]. The underdiagnosis of NAS observed in our rural cohort is consistent with previous reports [27, 28], which attribute this to limited antenatal screening [9] and fear of child removal [34], especially in First Nations communities [35], potentially delaying intervention and support [36].
Children with NAS are known to have increased morbidity and mortality in infancy and early childhood [5, 6, 14, 15, 37, 38]. Our study extends this to children with PDE, even without a diagnosis of NAS, and demonstrates this disparity persists into adolescence. The reasons for these adverse outcomes are multifactorial. Indeed, mothers in our PDE cohort were more likely to be younger [8, 9, 10], First Nations [5], lacking education [8, 10], unemployed [10], lower in socioeconomic status [5, 8, 9, 10] and more likely to live in unsafe, abusive circumstances [11], all aligning with prior observational studies and representing risk factors for childhood harm [39].
Temporal trends over the study period may offer insight into the efficacy of SUPS programs and rural funding initiatives, though whether services for children with PDE truly reach those in remote communities remains uncertain [40]. Our maps broadly demonstrate decreased rates of adverse outcomes for children with PDE in most metropolitan regions, while ED presentations increased in many regional and remote areas. The reasons for these findings cannot be speculated from administrative data and require further study.
Existing literature stratifying outcomes of PDE by remoteness is largely constrained to infancy and supports our finding of increased healthcare utilisation among rural infants with NAS [7, 33]. However, beyond infancy, regional children with PDE had lower rates of hospitalisation, ED presentations and placements in OOHC compared to metropolitan counterparts, yet were more likely to die. This novel finding may indicate that reduced healthcare utilisation in later childhood does not reflect lesser need in rural areas, but rather unmet need. It may also suggest that loss to follow‐up, a well‐recognised challenge in studying PDE populations that is inherently difficult to capture in longitudinal studies, disproportionately affects rural populations.
There are several strengths to our study. We demonstrate outcomes for children even without a diagnosis of PDE or NAS. Our extensive cohort allowed stratification of exposure groups, including to licit substances such as tobacco and alcohol [35].
This study has limitations. The probabilistic methods of data linkage carry a small risk of incorrect linkages (approximately 0.5%). Participants born later in the study period had less follow‐up information, and the depth of such follow‐up was highly dependent on the amount of contact with child protection and inpatient health services. Drug use may be under‐reported. Participants may not remain within their place of birth. The diagnosis of PDE may be under‐reported. While NAS is well recorded because these newborns receive a standardised treatment intervention, this may not be the case for PDE without NAS, as NAS usually refers to opiate withdrawal. This would lead to conservative reporting of effect sizes. There may be a cohort time effect over 20 years, which might not have been accurately adjusted for using calendar year. Interventions for PDE and children with PDE may have occurred at different times over different parts of the state. It is difficult to estimate the impact of the time‐cohort effect on these results.
6. Conclusion
Regional and remote areas are increasingly and disproportionately affected by PDE, where children use less health and social services but have higher mortality, even until 15 years of age. Despite government initiatives in Australia [29], geographical disparities persist among children with PDE and may be growing. This study emphasises the need to evaluate models of care to support women and children with PDE in regional and remote areas to prevent adverse outcomes beyond the newborn period.
Author Contributions
T.C. developed the project idea and methodology with Prof. Ju Lee Oei and Dr. Mithilesh Dronavalli, performed statistical analysis, drafted the initial manuscript, prepared and edited the final manuscript to be submitted. M.D. developed the project idea, merged and cleaned the datasets, provided statistical supervision to Ms. Colligan, assisted with statistical interpretation, revised and approved the final manuscript to be submitted. J.L.O. developed the project idea, obtained ethics approval, submitted the initial proposal to the Centre for Health and Record Linkage (CHeReL), reviewed and approved the final manuscript to be submitted, and provided editorial input on two drafts of this manuscript. J.L.O. is responsible for Indigenous oversight of the manuscript and the project. L.D. is the consumer representative. All authors (Taylor Colligan, Ju Lee Oei, Barbara Bajuk, Lucinda Burns, Kate Lawler, Hannah Uebel, John Eastwood, Andrew Page, Evelyn Lee, Lauren Dicair, Charles Green, Michelle Dickson, Mithilesh Dronavalli) approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Funding
This work was supported by SPHERE Mindgardens Neuroscience Network, Australian Red Cross, Alpha Maxx Healthcare, NHMRC (APP1198477) and University of Sydney.
Conflicts of Interest
The authors declare no conflicts of interest. Ju Lee Oei has received honoraria for speaking, travel and research from Mallinkrodt Inc.
Supporting information
Table S1: Summary of databases in record linkage.
Figure S1: Directed acyclic graph for relevant confounders in regression models.
Acknowledgements
Open access publishing facilitated by University of New South Wales, as part of the Wiley ‐ University of New South Wales agreement via the Council of Australasian University Librarians.
Data Availability Statement
The data that support the findings of this study are available from Ministry of Health, ChEREL. Restrictions apply to the availability of these data, which were used under license for this study. Data are available from https://www.cherel.org.au/ with the permission of Ministry of Health, ChEREL.
References
- 1. Allegaert K. and van den Anker J. N., “Neonatal Withdrawal Syndrome: Reaching Epidemic Proportions Across the Globe,” Archives of Disease in Childhood. Fetal and Neonatal Edition 101, no. 1 (2016): F2–F3. [DOI] [PubMed] [Google Scholar]
- 2. Al‐Enazy S., Ali S., Albekairi N., El‐Tawil M., and Rytting E., “Placental Control of Drug Delivery,” Advanced Drug Delivery Reviews 116 (2017): 63–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Coyle M. G., Brogly S. B., Ahmed M. S., Patrick S. W., and Jones H. E., “Neonatal Abstinence Syndrome,” Nature Reviews. Disease Primers 4, no. 1 (2018): 47. [DOI] [PubMed] [Google Scholar]
- 4. Perlstein M. A., “Congenital Morphinism; A Rare Cause of Convulsions in the Newborn,” Journal of the American Medical Association 135, no. 10 (1947): 633. [DOI] [PubMed] [Google Scholar]
- 5. Uebel H., Wright I. M., Burns L., et al., “Reasons for Rehospitalization in Children Who Had Neonatal Abstinence Syndrome,” Pediatrics 136, no. 4 (2015): e811–e820. [DOI] [PubMed] [Google Scholar]
- 6. Corr T. E., Xing X., and Liu G., “Longitudinal Health Care Utilization of Medicaid‐Insured Children With a History of Neonatal Abstinence Syndrome,” Journal of Pediatrics 233 (2021): 82–9.e1. [DOI] [PubMed] [Google Scholar]
- 7. Fang S.‐Y., Huang N., Lin T., Ho I.‐K., and Chen C.‐Y., “Health Insurance Coverage and Healthcare Utilization Among Infants of Mothers in the National Methadone Maintenance Treatment Program in Taiwan,” Drug and Alcohol Dependence 153 (2015): 86–93. [DOI] [PubMed] [Google Scholar]
- 8. Mravčík V., Nechanská B., Gabrhelík R., Handal M., Mahic M., and Skurtveit S., “Socioeconomic Characteristics of Women With Substance Use Disorder During Pregnancy and Neonatal Outcomes in Their Newborns: A National Registry Study From The Czech Republic,” Drug and Alcohol Dependence 209 (2020): 107933. [DOI] [PubMed] [Google Scholar]
- 9. Uebel H., Wright I. M., Burns L., et al., “Characteristics and Causes of Death in Children With Neonatal Abstinence Syndrome,” Journal of Paediatrics and Child Health 56, no. 12 (2020): 1933–1940. [DOI] [PubMed] [Google Scholar]
- 10. Witt C. E., Rudd K. E., Bhatraju P., Rivara F. P., Hawes S. E., and Weiss N. S., “Neonatal Abstinence Syndrome and Early Childhood Morbidity and Mortality in Washington State: A Retrospective Cohort Study,” Journal of Perinatology 37, no. 10 (2017): 1124–1129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Victor B. G., Grogan‐Kaylor A., Ryan J. P., Perron B. E., and Gilbert T. T., “Domestic Violence, Parental Substance Misuse and the Decision to Substantiate Child Maltreatment,” Child Abuse & Neglect 79 (2018): 31–41. [DOI] [PubMed] [Google Scholar]
- 12. Ko J. Y., Yoon J., Tong V. T., et al., “Maternal Opioid Exposure, Neonatal Abstinence Syndrome, and Infant Healthcare Utilization: A Retrospective Cohort Analysis,” Drug and Alcohol Dependence 223 (2021): 108704. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Shrestha S., Roberts M. H., Maxwell J. R., Leeman L. M., and Bakhireva L. N., “Post‐Discharge Healthcare Utilization in Infants With Neonatal Opioid Withdrawal Syndrome,” Neurotoxicology and Teratology 86 (2021): 106975. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Gabrhelík R., Skurtveit S., Nechanská B., Mravčík V., and Handal M., “Morbidity Through 3 Years of Age in Children of Women Using Methamphetamine During Pregnancy: A National Registry Study,” European Addiction Research 29, no. 1 (2023): 19–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Liu G., Kong L., Leslie D. L., and Corr T. E., “A Longitudinal Healthcare Use Profile of Children With a History of Neonatal Abstinence Syndrome,” Journal of Pediatrics 204 (2019): 111–7.e1. [DOI] [PubMed] [Google Scholar]
- 16. Brown J. D., Goodin A. J., and Talbert J. C., “Rural and Appalachian Disparities in Neonatal Abstinence Syndrome Incidence and Access to Opioid Abuse Treatment,” Journal of Rural Health 34, no. 1 (2018): 6–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Shaw M. R., Grant T., Barbosa‐Leiker C., Fleming S. E., Henley S., and Graham J. C., “Intervention With Substance‐Abusing Mothers: Are There Rural‐Urban Differences?,” American Journal on Addictions 24, no. 2 (2015): 144–152. [DOI] [PubMed] [Google Scholar]
- 18. Campbell F. A., Pungello E. P., Burchinal M., et al., “Adult Outcomes as a Function of an Early Childhood Educational Program: An Abecedarian Project Follow‐Up,” Developmental Psychology 48, no. 4 (2012): 1033–1043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Australian Institute of Health Welfare , “Alcohol and Other Drug Use in Regional and Remote Australia: Consumption, Harms and Access to Treatment 2016–17,” 2019, https://www.aihw.gov.au/reports/alcohol‐other‐drug‐treatment‐services/alcohol‐other‐drug‐use‐regional‐remote‐2016‐17/contents/summary.
- 20. Kozhimannil K. B., Chantarat T., Ecklund A. M., Henning‐Smith C., and Jones C., “Maternal Opioid Use Disorder and Neonatal Abstinence Syndrome Among Rural US Residents, 2007–2014,” Journal of Rural Health 35, no. 1 (2019): 122–132. [DOI] [PubMed] [Google Scholar]
- 21. Harris D. E., Aboueissa A. M., Baugh N., and Sarton C., “Impact of Rurality on Maternal and Infant Health Indicators and Outcomes in Maine,” Rural and Remote Health 15, no. 3 (2015): 3278. [PubMed] [Google Scholar]
- 22. Aljassim N. and Ostini R., “Health Literacy in Rural and Urban Populations: A Systematic Review,” Patient Education and Counseling 103, no. 10 (2020): 2142–2154. [DOI] [PubMed] [Google Scholar]
- 23. Chen X., Orom H., Hay J. L., et al., “Differences in Rural and Urban Health Information Access and Use,” Journal of Rural Health 35, no. 3 (2019): 405–417. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Warwick S., Atkinson D., Kitaura T., LeLievre M., and Marley J. V., “Young Aboriginal People's Perspective on Access to Health Care in Remote Australia: Hearing Their Voices,” Progress in Community Health Partnerships: Research, Education, and Action 13, no. 2 (2019): 171–181. [DOI] [PubMed] [Google Scholar]
- 25. Rodrigues A. N., Bajuk B., Oei J., Lui K., and Abdel‐Latif M. E., “Neurodevelopmental Outcome of Extremely Preterm Infants Born to Rural and Urban Residents' Mothers in Australia,” Early Human Development 91, no. 8 (2015): 437–443. [DOI] [PubMed] [Google Scholar]
- 26. Baxter J., Hayes A., and Gray M., Families in Regional, Rural and Remote (Australian Institute of Family Studies, 2011). [Google Scholar]
- 27. Huizink A. C., “Prenatal Cannabis Exposure and Infant Outcomes: Overview of Studies,” Progress in Neuro‐Psychopharmacology & Biological Psychiatry 52 (2014): 45–52. [DOI] [PubMed] [Google Scholar]
- 28. Jones H. E., Kaltenbach K., Benjamin T., Wachman E. M., and O'Grady K. E., “Prenatal Opioid Exposure, Neonatal Abstinence Syndrome/Neonatal Opioid Withdrawal Syndrome, and Later Child Development Research: Shortcomings and Solutions,” Journal of Addiction Medicine 13, no. 2 (2019): 90–92. [DOI] [PubMed] [Google Scholar]
- 29. Australian Government Department of Health , “Budget 2022–23: Building Upon the Stronger Rural Health Strategy,” 2022.
- 30. Centre for Health Record Linkage , “Our Services—CHeReL,” 2023, https://www.cherel.org.au/how‐record‐linkage‐works.
- 31. Australian Bureau of Statistics , “Remoteness Areas,” 2023, https://www.abs.gov.au/statistics/standards/australian‐statistical‐geography‐standard‐asgs‐edition‐3/jul2021‐jun2026/remoteness‐structure/remoteness‐areas.
- 32. Pisati M., “SPMAP: Stata Module to Visualize Spatial Data,” 2018.
- 33. Salt E., Wiggins A., Pick A., et al., “A Case‐Control Study Comparing Rates and Diagnoses of Hospital Readmission in Infants Affected by Neonatal Abstinence Syndrome,” Journal of Maternal‐Fetal and Neonatal Medicine 36, no. 1 (2023): 2162820. [DOI] [PubMed] [Google Scholar]
- 34. Abdel‐Latif M. E., Oei J., Craig F., et al., “Profile of Infants Born to Drug‐Using Mothers: A State‐Wide Audit,” Journal of Paediatrics and Child Health 49, no. 1 (2013): E80–E86. [DOI] [PubMed] [Google Scholar]
- 35. O'Connor A., Harris E., Seeber C., Hamilton D., Fisher C., and Sachmann M., “Methamphetamine Use in Pregnancy, Child Protection, and Removal of Infants: Tertiary Centre Experience From Western Australia,” Midwifery 83 (2020): 102641. [DOI] [PubMed] [Google Scholar]
- 36. Leyenaar J. K., Schaefer A. P., Wasserman J. R., Moen E. L., O'Malley A. J., and Goodman D. C., “Infant Mortality Associated With Prenatal Opioid Exposure,” JAMA Pediatrics 175, no. 7 (2021): 706–714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Lisonkova S., Richter L. L., Ting J., et al., “Neonatal Abstinence Syndrome and Associated Neonatal and Maternal Mortality and Morbidity,” Pediatrics 144, no. 2 (2019): e20183664. [DOI] [PubMed] [Google Scholar]
- 38. Gaither J. R., Drago M. J., Grossman M. R., et al., “Hospital Readmissions Among Infants With Neonatal Opioid Withdrawal Syndrome,” JAMA Network Open 7, no. 9 (2024): e2435074‐e. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Chen D., Lin L., Li C., et al., “Maternal Adverse Childhood Experiences and Health‐Related Quality of Life in Preschool Children: A Cross‐Sectional Study,” Child and Adolescent Psychiatry and Mental Health 17, no. 1 (2023): 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Coupland H., Moensted M. L., Reid S., et al., “Developing a Model of Care for Substance Use in Pregnancy and Parenting Services, Sydney, Australia: Service Provider Perspectives,” Journal of Substance Abuse Treatment 131 (2021): 108420. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Summary of databases in record linkage.
Figure S1: Directed acyclic graph for relevant confounders in regression models.
Data Availability Statement
The data that support the findings of this study are available from Ministry of Health, ChEREL. Restrictions apply to the availability of these data, which were used under license for this study. Data are available from https://www.cherel.org.au/ with the permission of Ministry of Health, ChEREL.