Prediction of pre-eclampsia with machine learning approaches: Leveraging important information from routinely collected data.

BACKGROUND: Globally, pre-eclampsia (PE) is a leading cause of maternal and perinatal morbidity and mortality. PE prediction using routinely collected data has the advantage of being widely applicable, particularly in low-resource settings. Early intervention for high-risk women might reduce PE incidence and related complications. We aimed to replicate our machine learning (ML) published work predicting another maternal condition (gestational diabetes) to (1) predict PE using routine health data, (2) identify the optimal ML model, and (3) compare it with logistic regression approach. METHODS: Data were from a large health service network with 48,250 singleton pregnancies between January 2016 and June 2021. Supervised ML models were employed. Maternal clinical and medical characteristics were the feature variables (predictors), and a 70/30 data split was used for training and testing the model. Predictive performance was assessed using area under the curve (AUC) and calibration plots. Shapley value analysis assessed the contribution of feature variables. RESULTS: The random forest approach provided excellent discrimination with an AUC of 0.84 (95% CI: 0.82-0.86) and highest prediction accuracy (0.79); however, the calibration curve (slope of 1.21, 95% CI 1.13-1.30) was acceptable only for a threshold of 0.3 or less. The next best approach was extreme gradient boosting, which provided an AUC of 0.77 (95% CI: 0.76-0.79) and well-calibrated (slope of 0.93, 95% CI 0.85-1.01). Logistic regression provided good discrimination performance with an AUC of 0.75 (95% CI: 0.74-0.76) and perfect calibration. Nulliparous, pre-pregnancy body mass index, previous pregnancy with prior PE, maternal age, family history of hypertension, and pre-existing hypertension and diabetes were the top-ranked features in Shapley value analysis. CONCLUSION: Two ML models created the highest-performing prediction using routinely collected data to identify women at high risk of PE, with acceptable discrimination. However, to confirm this result and also examine model generalisability, external validation studies are needed in other settings, utilising standardised prognostic factors.

The Use of Multiple Medications During Pregnancy Among an Ethnically Diverse Population in South-Eastern Melbourne: A Retrospective Analysis to Explore Potential Risks and Complications.

BACKGROUND AND OBJECTIVE: Medication use is increasing to treat both pre-existing and pregnancy-related medical conditions or complications. This study aims to investigate factors associated with multiple medication use during pregnancy, as well as any increased risk of pregnancy complications for women taking multiple medications. METHODS: A retrospective analysis of routinely collected medical records of singleton pregnant women was conducted in Southeast Melbourne, Australia, between 2016 and 2021. Self-reported medication use was recorded as part of routine medical care, starting from the first antenatal booking appointment and continuing for every subsequent antenatal appointment until birth. Multimorbidity was defined as having two or more medical conditions. Logistic regression was used to assess factors influencing multiple medication use (defined as taking two or more non-supplemental medications at any stage of pregnancy) and associations with pregnancy complications. RESULTS: Of 48,502 participants, 34.9% used one medication, while 11.7% used multiple medications. Women of older age (30-34, 35-39, and ≥  40 years), higher body mass index (25.0-29.9 kg/m2 and ≥  30 kg/m2), born in Australasia and Oceania, higher socioeconomic status, and multimorbidity were more likely to use multiple medications during pregnancy. Women taking multiple medications had a higher risk of preterm and caesarean deliveries, fetal death, and neonatal admissions to intensive care. Sensitivity analyses exploring different morbidity categories produced no changes to findings. CONCLUSIONS: Medication use during pregnancy is prevalent, with many pregnant mothers taking multiple medications. Given the rising maternal age, body mass index, and morbidities in pregnancy, the use of medications during pregnancy is increasing. Such use correlates with an increased chance of adverse pregnancy outcomes. In the context of limited trials on the safety and efficacy of medications in pregnancy, timely harnessing of the information available within routine medical records for post-marketing surveillance is important.

Tools and frameworks for evaluating the implementation of learning health systems: a scoping review.

INTRODUCTION: Despite increased interest in learning health systems (LHS), a paucity of guidance and tools for evaluating LHS implementation exists. To address this, we aim to undertake a scoping review on existing tools and evaluation of exemplars of LHS implementation. METHODS: We conducted a scoping review of peer-reviewed studies within Scopus, EMBASE, MEDLINE, and MEDLINE in-process that described (1) the evaluation of the implementation of an operating LHS or (2) the development of a framework or tool to facilitate this evaluation. Anima, basic research, abstracts, non-English language articles, and publications before 2018 were excluded. All study designs were considered. FINDINGS: From 1300 studies initially identified, 4 were eligible, revealing three tools with nine implementation evaluation examples. The identified tools shared constructs which were evaluated, including: Stakeholders, Data, Research Evidence, Implementation, and Sociotechnical Infrastructure. However, there was divergence in evaluation methodology. Tools ranged from a five-point numerical rating system for process maturity with a radar chart called the Network Maturity Grid (NMG); the Kaiser Permanente Washington (KPWA) LHS Logic Model, which provides a broad list of constructs and sample measures relevant to LHS operations; and finally LADDERS, a simple tool or form-based template designed for consistent evaluation over time. The NMG tool was the most mature in terms of adaptation and adoption. Notably, two (NMG and the KPWA LHS Logic Model) out of three tools conceptualized the LHS as a suite of processes and devised tools were processes that linked these constructs. IMPLICATIONS FOR TOOLKIT DEVELOPMENT: The evaluation of LHS implementation remains an under explored area of investigation, as this scoping review found only three tools for LHS implementation evaluation. Our findings indicate a need for further empirical research in this area and suggest early consensus in constructs that need to be considered during evaluation.

Relationship Between Early-Pregnancy Glycemia and Adverse Outcomes: Findings From the TOBOGM Study.

OBJECTIVE: We evaluated associations between early-pregnancy oral glucose tolerance test (OGTT) glucose and complications in the Treatment of Booking Gestational Diabetes Mellitus (TOBOGM) cohort to inform prognostic OGTT thresholds. RESEARCH DESIGN AND METHODS: Individuals with risk factors for hyperglycemia were recruited for an international, multicenter, randomized controlled gestational diabetes mellitus (GDM) (World Health Organization 2013 criteria) treatment trial. A 2-h 75-g OGTT was performed at <20 weeks' gestation. Individuals with early treated hyperglycemia in pregnancy were excluded from the primary analysis. Early OGTT glucose concentrations were analyzed continuously and in glycemic categories (normal, low band, and high band). RESULTS: Overall, 3,645 individuals had an OGTT at (mean ± SD) 15.6 ± 2.5 weeks. For each 1-SD increase in fasting, 1-h, and 2-h glucose values, there were continuous positive associations with late GDM: adjusted odds ratio (aOR) 2.04 (95% CI 1.82-2.27), 3.05 (2.72-3.43), and 2.21 (1.99-2.45), respectively. There were continuous positive associations between 1-h and 2-h glucose and the perinatal composite (birth <37 + 0 weeks, birth trauma, birth weight ≥4,500 g, respiratory distress, phototherapy requirement, stillbirth/neonatal death, and shoulder dystocia), with aOR 1.15 (95% CI 1.04-1.26) and 1.14 (1.04-1.25), respectively, and with large-for-gestational-age offspring, with aOR 1.18 (1.06-1.31) and 1.26 (1.01-1.25), respectively. Significant associations were also observed between 1-h glucose and cesarean section and between fasting and 2-h glucose and neonatal hypoglycemia. In categorical analysis, only the high-band 1-h glucose (≥10.6 mmol/L [191 mg/dL]) predicted the perinatal composite. CONCLUSIONS: There is a continuous positive association between early-pregnancy OGTT glucose and complications. In individuals with hyperglycemia risk factors, only the high-glycemic-band 1-h glucose corresponded to increased risk of major perinatal complications.

External validation of risk prediction model for gestational diabetes: Individual participant data meta-analysis of randomized trials.

BACKGROUND: An original validated risk prediction model with good discriminatory prognostic performance for predicting gestational diabetes (GDM) diagnosis, has been updated for recent international association of diabetes in pregnancy study group (IADPSG) diagnostic criteria. However, the updated model is yet to be externally validated on an international dataset. AIMS: To perform an external validation of the updated risk prediction model to evaluate model indices such as discrimination and calibration based on data from the International Weight Management in Pregnancy (i-WIP) Collaborative Group. MATERIALS AND METHODS: The i -WIP dataset was used to validate the GDM prediction tool across discrimination and model calibration. RESULTS: Overall 7689 individual patient data were included, with 17.4 % with GDM, however only 113 cases were available using IADPSG (International Association of Diabetes and Pregnancy Groups) criteria for 75 g OGTT glucose load and ACOG (American College of Obstetricians and Gynecologists) for 100 g glucose load and having the routine clinical risk factor data. The GDM model was moderately discriminatory (Area Under the Curve (AUC) of 0.67; 95 % CI 0.59 to 0.75), Sensitivity 81.0 % (95 % CI 66.7 % to 90.9 %), specificity 53 % (40.3 % to 65.4 %). The GDM score showed reasonable calibration for predicting GDM (slope = 0.84, CITL = 0.77). Imputation for missing data increased the sample to n = 253, and vastly improved the discrimination and calibration of the model to AUC = 78 (95 % CI 72 to 85), sensitivity (81 %, 95 % CI 66.7 % to 90.9 %) and specificity (75 %, 95 % CI 68.8 % to 81 %). CONCLUSION: The updated GDM model showed promising discrimination in predicting GDM in an international population sourced from RCT individual patient data. External validations are essential in order for the risk prediction area to advance, and we demonstrate the utility of using existing RCT data from different global settings. Despite limitations associated with harmonising the data to the variable types in the model, the validation model indices were reasonable, supporting generalizability across continents and populations.

Identifying Predictor Variables for a Composite Risk Prediction Tool for Gestational Diabetes and Hypertensive Disorders of Pregnancy: A Modified Delphi Study.

A composite cardiometabolic risk prediction tool will support the systematic identification of women at increased cardiometabolic risk during pregnancy to enable early screening and intervention. This study aims to identify and select predictor variables for a composite risk prediction tool for cardiometabolic risk (gestational diabetes mellitus and/or hypertensive disorders of pregnancy) for use in the first trimester. A two-round modified online Delphi study was undertaken. A prior systematic literature review generated fifteen potential predictor variables for inclusion in the tool. Multidisciplinary experts (n = 31) rated the clinical importance of variables in an online survey and nominated additional variables for consideration (Round One). An online meeting (n = 14) was held to deliberate the importance, feasibility and acceptability of collecting variables in early pregnancy. Consensus was reached in a second online survey (Round Two). Overall, 24 variables were considered; 9 were eliminated, and 15 were selected for inclusion in the tool. The final 15 predictor variables related to maternal demographics (age, ethnicity/race), pre-pregnancy history (body mass index, height, history of chronic kidney disease/polycystic ovarian syndrome, family history of diabetes, pre-existing diabetes/hypertension), obstetric history (parity, history of macrosomia/pre-eclampsia/gestational diabetes mellitus), biochemical measures (blood glucose levels), hemodynamic measures (systolic blood pressure). Variables will inform the development of a cardiometabolic risk prediction tool in subsequent research. Evidence-based, clinically relevant and routinely collected variables were selected for a composite cardiometabolic risk prediction tool for early pregnancy.

Geographic Variation in Osteoporosis Treatment in Postmenopausal Women: A 15-Year Longitudinal Analysis.

Context: Osteoporosis affects more than half of older women, but many are not treated. Whether treatment differs between rural and urban areas is unknown. Objective: To examine differences in osteoporosis treatment among postmenopausal women living in urban and rural areas of Australia. Methods: Women participating in the Australian Longitudinal Study on Women's Health, a prospective longitudinal cohort study, born between 1946-1951, and with osteoporosis or fractures, were included. Surveys from 2004 to 2019 were linked to the Pharmaceutical Benefits Scheme (government-subsidized medications) to assess osteoporosis treatment and adherence, comparing geographical areas. Results: Of the 4259 women included (mean age, 55.6 years), 1703 lived in major cities, 1629 inner regional, 794 outer regional, and 133 remote areas. Over the 15-year follow-up, 1401 (32.9%) women received treatment, including 47.4% of women with osteoporosis and 29.9% with fractures. Women in outer regional and remote areas were less likely to use antiosteoporosis treatment than those in major cities on univariable analysis (outer regional odds ratio, 0.83; 95% CI, 0.72-0.95; remote, 0.65; 0.49-0.86), but this did not remain significant on multivariable analysis. Median duration of use was 10 to 36 months, adherence varied by treatment type (34%-100%) but was not related to incident fractures, and of the women who stopped denosumab, 85% did not receive another consolidating treatment. Conclusions: One-third of women with osteoporosis/fractures received treatment, and adherence was low. There was no difference in treatment use between urban and rural areas after adjusting for risk factors, although the specific treatment used, and adherence, differed.

Comparison of parent-reported motivators of non-vaccination for children 5-11 years old in Australia and Canada: Results of the iCARE study.

BACKGROUND: Despite COVID-19 infection being less severe in children compared to adults, vaccination for children from the age of 6 months onwards is recommended in many countries to reduce symptom severity and prevent severe disease. However, vaccination against COVID-19 for children remains controversial and uptake has been low. AIMS: To assess and compare the rate of change of parent-reported COVID-19 vaccine uptake in children aged 5 to 11 years and motivators of vaccine acceptance and non-vaccination among parents/guardians in Canada and Australia. METHODS: As part of the iCARE study, two cross-sectional representative samples in Canada and Australia were collected between May 20 and September 12, 2022 (i.e., 5 and 9 months after the COVID-19 vaccine rollout for children 5-11 years) using online panels. Parents/guardians reported the vaccine status of their children and motivators for vaccine acceptance and non-vaccination. General linear models were used to estimate differences between countries in terms of vaccine uptake and motivators across time. RESULTS: Parent-reported vaccine uptake for children 5-11 years didn't increase over the study period (T1 = 87 %,T2 = 86 %; OR = 0.83; 95 %CI = 0.45-1.54) and was overall lower in Canada (60.8 %) compared to Australia (71.6 %)(OR = 0.56; 95 %CI = 0.33-0.96). In both countries the socioeconomic characteristics of parents who didn't vaccinate their children were similar and having information on either the short- or long-term side effects of the vaccine were important motivators. However, vaccine effectiveness was more important in Canada and trust in the company that developed the vaccine and a recommendation from the child's doctor were more important motivators in Australia. CONCLUSION: Parent-reported vaccine uptake for children 5-11 years plateaued early in the vaccine rollout. The main motivators for parents of unvaccinated children varied between the two countries but information on vaccine safety and effectiveness were common to both countries. Findings may inform future tailored vaccine communication efforts and pandemic planning in Australia and Canada to optimize vaccine uptake for primary school children.

Development of a risk prediction model for postpartum onset of type 2 diabetes mellitus, following gestational diabetes; the lifestyle InterVention in gestational diabetes (LIVING) study.

AIMS: This study aimed to develop a prediction model for identifying a woman with gestational diabetes mellitus (GDM) at high risk of type 2 diabetes (T2DM) post-birth. METHODS: Utilising data from 1299 women in the Lifestyle Intervention IN Gestational Diabetes (LIVING) study, two models were developed: one for pregnancy and another for postpartum. Key predictors included glucose test results, medical history, and biometric indicators. RESULTS: Of the initial cohort, 124 women developed T2DM within three years. The study identified seven predictors for the antenatal T2DM risk prediction model and four for the postnatal one. The models demonstrated good to excellent predictive ability, with Area under the ROC Curve (AUC) values of 0.76 (95% CI: 0.72 to 0.80) and 0.85 (95% CI: 0.81 to 0.88) for the antenatal and postnatal models, respectively. Both models underwent rigorous validation, showing minimal optimism in predictive capability. Antenatal model, considering the Youden index optimal cut-off point of 0.096, sensitivity, specificity, and accuracy were measured as 70.97%, 70.81%, and 70.82%, respectively. For the postnatal model, considering the cut-off point 0.086, sensitivity, specificity, and accuracy were measured as 81.40%, 75.60%, and 76.10%, respectively. CONCLUSIONS: These models are effective for predicting T2DM risk in women with GDM, although external validation is recommended before widespread application.

Cost-effectiveness of diagnosis and treatment of early gestational diabetes mellitus: economic evaluation of the TOBOGM study, an international multicenter randomized controlled trial.

Background: A recently undertaken multicenter randomized controlled trial (RCT) "Treatment Of BOoking Gestational diabetes Mellitus" (TOBOGM: 2017-2022) found that the diagnosis and treatment of pregnant women with early gestational diabetes mellitus (GDM) improved pregnancy outcomes. Based on data from the trial, this study aimed to assess the cost-effectiveness of diagnosis and treatment of early GDM (from <20 weeks') among women with risk factors for hyperglycemia in pregnancy compared with usual care (no treatment until 24-28 weeks') from a healthcare perspective. Methods: Participants' healthcare resource utilization data were collected from their self-reported questionnaires and hospital records, and valued using the unit costs obtained from standard Australian national sources. Costs were reported in US dollars ($) using the purchasing power parity (PPP) estimates to facilitate comparison of costs across countries. Intention-to-treat (ITT) principle was followed. Missing cost data were replaced using multiple imputations. Bootstrapping method was used to estimate the uncertainty around mean cost difference and cost-effectiveness results. Bootstrapped cost-effect pairs were used to plot the cost-effectiveness (CE) plane and cost-effectiveness acceptability curve (CEAC). Findings: Diagnosis and treatment of early GDM was more effective and tended to be less costly, i.e., dominant (cost-saving) [-5.6% composite adverse pregnancy outcome (95% CI: -10.1%, -1.2%), -$1373 (95% CI: -$3,749, $642)] compared with usual care. Our findings were confirmed by both the CE plane (88% of the bootstrapped cost-effect pairs fall in the south-west quadrant), and CEAC (the probability of the intervention being cost-effective ranged from 84% at a willingness-to-pay (WTP) threshold value of $10,000-99% at a WTP threshold value of $100,000 per composite adverse pregnancy outcome prevented). Sub-group analyses demonstrated that diagnosis and treatment of early GDM among women in the higher glycemic range (fasting blood glucose 95-109 mg/dl [5.3-6.0 mmol/L], 1-h blood glucose ≥191 mg/dl [10.6 mmol/L] and/or 2-h blood glucose 162-199 mg/dl [9.0-11.0 mmol/L]) was more effective and less costly (dominant) [-7.8% composite adverse pregnancy outcome (95% CI: -14.6%, -0.9%), -$2795 (95% CI: -$6,638, -$533)]; the intervention was more effective and tended to be less costly [-8.9% composite adverse pregnancy outcome (95% CI: -15.1%, -2.6%), -$5548 (95% CI: -$16,740, $1547)] among women diagnosed before 14 weeks' gestation as well. Interpretation: Our findings highlight the potential health and economic benefits from the diagnosis and treatment of early GDM among women with risk factors for hyperglycemia in pregnancy and supports its implementation. Long-term follow-up studies are recommended as a key future area of research to assess the potential long-term health benefits and economic consequences of the intervention. Funding: National Health and Medical Research Council (grants 1104231 and 2009326), Region O¨rebro Research Committee (grants Dnr OLL-970566 and OLL-942177), Medical Scientific Fund of the Mayor of Vienna (project 15,205 and project 23,026), South Western Sydney Local Health District Academic Unit (grant 2016), and Western Sydney University Ainsworth Trust Grant (2019).

Machine Learning Algorithms Versus Classical Regression Models in Pre-Eclampsia Prediction: A Systematic Review.

PURPOSE OF REVIEW: Machine learning (ML) approaches are an emerging alternative for healthcare risk prediction. We aimed to synthesise the literature on ML and classical regression studies exploring potential prognostic factors and to compare prediction performance for pre-eclampsia. RECENT FINDINGS: From 9382 studies retrieved, 82 were included. Sixty-six publications exclusively reported eighty-four classical regression models to predict variable timing of onset of pre-eclampsia. Another six publications reported purely ML algorithms, whilst another 10 publications reported ML algorithms and classical regression models in the same sample with 8 of 10 findings that ML algorithms outperformed classical regression models. The most frequent prognostic factors were age, pre-pregnancy body mass index, chronic medical conditions, parity, prior history of pre-eclampsia, mean arterial pressure, uterine artery pulsatility index, placental growth factor, and pregnancy-associated plasma protein A. Top performing ML algorithms were random forest (area under the curve (AUC) = 0.94, 95% confidence interval (CI) 0.91-0.96) and extreme gradient boosting (AUC = 0.92, 95% CI 0.90-0.94). The competing risk model had similar performance (AUC = 0.92, 95% CI 0.91-0.92) compared with a neural network. Calibration performance was not reported in the majority of publications. ML algorithms had better performance compared to classical regression models in pre-eclampsia prediction. Random forest and boosting-type algorithms had the best prediction performance. Further research should focus on comparing ML algorithms to classical regression models using the same samples and evaluation metrics to gain insight into their performance. External validation of ML algorithms is warranted to gain insights into their generalisability.

The impact of elder abuse training on subacute health providers and older adults: study protocol for a randomized control trial.

BACKGROUND: Elder abuse often goes unreported and undetected. Older people may be ashamed, fearful, or otherwise reticent to disclose abuse, and many health providers are not confident in asking about it. In the No More Shame study, we will evaluate a co-designed, multi-component intervention that aims to improve health providers' recognition, response, and referral of elder abuse. METHODS: This is a single-blinded, pragmatic, cluster randomised controlled trial. Ten subacute hospital sites (i.e. clusters) across Australia will be allocated 1:1, stratified by state to a multi-component intervention comprising a training programme for health providers, implementation of a screening tool and use of site champions, or no additional training or support. Outcomes will be collected at baseline, 4 and 9 months. Our co-primary outcomes are change in health providers' knowledge of responding to elder abuse and older people's sense of safety and quality of life. We will include all inpatients at participating sites, aged 65 + (or aged 50 + if Aboriginal or Torres Strait Islander), who are able to provide informed consent and all unit staff who provide direct care to older people; a sample size of at least 92 health providers and 612 older people will provide sufficient power for primary analyses. DISCUSSION: This will be one of the first trials in the world to evaluate a multi-component elder abuse intervention. If successful, it will provide the most robust evidence base to date for health providers to draw on to create a safe environment for reporting, response, and referral. TRIAL REGISTRATION: ANZCTR, ACTRN12623000676617p . Registered 22 June 2023.

Learning together for better health using an evidence-based Learning Health System framework: a case study in stroke.

BACKGROUND: In the context of expanding digital health tools, the health system is ready for Learning Health System (LHS) models. These models, with proper governance and stakeholder engagement, enable the integration of digital infrastructure to provide feedback to all relevant parties including clinicians and consumers on performance against best practice standards, as well as fostering innovation and aligning healthcare with patient needs. The LHS literature primarily includes opinion or consensus-based frameworks and lacks validation or evidence of benefit. Our aim was to outline a rigorously codesigned, evidence-based LHS framework and present a national case study of an LHS-aligned national stroke program that has delivered clinical benefit. MAIN TEXT: Current core components of a LHS involve capturing evidence from communities and stakeholders (quadrant 1), integrating evidence from research findings (quadrant 2), leveraging evidence from data and practice (quadrant 3), and generating evidence from implementation (quadrant 4) for iterative system-level improvement. The Australian Stroke program was selected as the case study as it provides an exemplar of how an iterative LHS works in practice at a national level encompassing and integrating evidence from all four LHS quadrants. Using this case study, we demonstrate how to apply evidence-based processes to healthcare improvement and embed real-world research for optimising healthcare improvement. We emphasize the transition from research as an endpoint, to research as an enabler and a solution for impact in healthcare improvement. CONCLUSIONS: The Australian Stroke program has nationally improved stroke care since 2007, showcasing the value of integrated LHS-aligned approaches for tangible impact on outcomes. This LHS case study is a practical example for other health conditions and settings to follow suit.

A longitudinal study of the impacts of a stay in a Prevention and Recovery Care service in Victoria, Australia.

BACKGROUND: Prevention and Recovery Care services are residential sub-acute services in Victoria, Australia, guided by a commitment to recovery-oriented practice. The evidence regarding the effectiveness of this service model is limited, largely relying on small, localised evaluations. This study involved a state-wide investigation into the personal recovery, perceived needs for care, well-being and quality-of-life outcomes experienced by Prevention and Recovery Care services' consumers. METHODS: A longitudinal cohort design examined the trajectory of self-reported personal recovery and other outcomes for consumers in 19 Victorian Prevention and Recovery Care services over 4 time points (T1 - 1 week after admission; T2 - within 1 week of discharge; T3 - 6 months after discharge; T4 - 12 months after discharge). T2-T4 time frames were extended by approximately 3 weeks due to recruitment challenges. The Questionnaire about the Process of Recovery was the primary outcome measure. RESULTS: At T1, 298 consumers were recruited. By T4, 114 remained in the study. Participants scored higher on the Questionnaire about the Process of Recovery at all three time points after T1. There were also sustained improvements on all secondary outcome measures. Improvements were then sustained at each subsequent post-intervention time point. Community inclusion and having needs for care met also improved. CONCLUSION: The findings provide a consistent picture of benefits for consumers using Prevention and Recovery Care services, with significant improvement in personal recovery, quality of life, mental health and well-being following an admission to a Prevention and Recovery Care service. Further attention needs to be given to how to sustain the gains made through a Prevention and Recovery Care service admission in the long term.

Trends and determinants of mental illness in humanitarian migrants resettled in Australia: Analysis of longitudinal data.

Given the frequent exposure of humanitarian migrants to traumatic or stressful circumstances, there exists a potential predisposition to mental illness. Our objective was to pinpoint the trends and determinants of mental illness among humanitarian migrants resettled in Australia. This study considered five waves of longitudinal data involving humanitarian migrants resettled in Australia. Post-traumatic stress disorder (PTSD) and psychological distress were assessed using PTSD-8 and Kessler-6 screening tools. Through a Generalised Linear Mixed model (GLMM), variables displaying a 95% CI that excluded the value of 1.0 for the odds ratio were identified as associated factors for both PTSD and elevated psychological distress. The selection of multivariable covariates was guided by causal loop diagrams and least absolute shrinkage and selection operators methods. At baseline, there were 2399 humanitarian migrants with 1881 retained and at the fifth yearly wave; the response rate was 78.4%. PTSD prevalence decreased from 33.3% (95% CI: 31.4-35.3) at baseline to 28.3% (95% CI: 26.2-30.5) at year 5. Elevated psychological distress persisted across all waves: 17.1% (95% CI: 15.5-18.6) at baseline and 17.1% (95% CI: 15.3-18.9) at year 5. Across the five waves, 34.0% of humanitarian migrants met screening criteria for mental illness, either PTSD or elevated psychological distress. In the multivariate model, factors associated with PTSD were loneliness (AOR 1.5, 95% CI: 1.3-1.8), discrimination (AOR 1.6: 1.2-2.1), temporary housing contract (AOR 3.7: 2.1-6.7), financial hardship (AOR 2.2:1.4-3.6) and chronic health conditions (AOR 1.3: 1.1-1.5), whereas the associated factors for elevated psychological distress were loneliness (AOR 1.8: 1.5-2.2), discrimination (AOR 1.7: 1.3-2.2) and short-term lease housing (AOR 1.6: 1.0-1.7). The prevalence, persistence and consequential burden of mental illness within this demographic underscore the urgent need for targeted social and healthcare policies. These policies should aim to mitigate modifiable risk factors, thereby alleviating the significant impact of mental health challenges on this population.

Diagnostic accuracy of oral glucose tolerance tests, fasting plasma glucose and haemoglobin A1c for type 2 diabetes in women with polycystic ovary syndrome: A systematic review and meta-analysis.

AIMS: To inform international guidelines, a systematic review and meta-analysis was conducted to assess the performance of diagnostic methods for type 2 diabetes in women with polycystic ovary syndrome (PCOS). METHODS: An updated systematic search was conducted on five databases from 2017 until October 2023 and combined with prior searches (from inception). Meta-analyses of diagnostic accuracy tests were conducted. RESULTS: Nine studies comprising 2628 women with PCOS were included. Against the oral glucose tolerance test, a haemoglobin A1C (HbA1c) ≥ 6.5% had a pooled sensitivity of 50.00% (95% confidence interval (CI): 35.53-64.47), specificity of 99.86% (95%CI: 99.49-99.98), and positive and negative predictive values of 92.59% (95%CI: 75.27-98.09) and 98.27% (95%CI: 97.73-98.68), respectively, with an accuracy of 98.17% (95%CI: 97.34-98.79). Fasting plasma glucose values ≥ 7.0 mmol/L had a pooled sensitivity of 58.14% (95%CI: 42.13-72.99), specificity of 92.59% (95%CI: 75.35-98.08), positive and negative predictive values of 92.59% (95%CI: 75.35-98.08) and 99.09% (95%CI: 98.71-99.36), respectively, and an accuracy of 99.00% (95%CI: 98.46-99.39) against the oral glucose tolerance test. CONCLUSIONS: To our knowledge, this is the first systematic review assessing the performance of diagnostic methods for type 2 diabetes in women with PCOS. We demonstrate that using a cut-off for HbA1c of ≥6.5% in this population may result in misdiagnosis of half of the women with type 2 diabetes. Our results directly informed the recommendations of the 2023 International PCOS Guideline, suggesting that the oral glucose tolerance test is the optimal method for screening and diagnosing type 2 diabetes in women with PCOS and is superior to fasting plasma glucose and HbA1c.

How the structural determinants of health inequities impact access to prescription medication for pregnant women in Australia: a narrative review.

Structural factors that contribute to health disparities (e.g., population-level policies, cultural norms) impact the distribution of resources in society and can affect medication accessibility; even in high-income countries like Australia. Industry practices and regulatory approaches (e.g., a conservative approach to testing medicines in pregnant women) influence the availability of safety and efficacy data necessary for the licencing and funding of prescription medications used during pregnancy. Consequently, pregnant women may be prescribed medications outside of regulatory or funder-approved indications, posing risks for both prescribers and pregnant women and potentially compromising equitable access to medications. This review examines the regulatory and legislative structural factors that contribute to health disparities and perpetuate the deeply ingrained social norm that we should be protecting pregnant women from clinical research rather than safeguarding them through such research. Addressing these challenges requires a renewed commitment to integrated, woman-centred maternal healthcare and strengthened collaboration across all sectors. Funding: Australian Government Research Training Program Stipend from the University of Technology Sydney, National Health and Medical Research Council (NHMRC) Fellowship, Channel 7 Children's Research Foundation Fellowship (CRF-210323).

Likelihood of primary cesarean section following induction of labor in singleton cephalic pregnancies at term, compared with expectant management: An Australian population-based, historical cohort study.

INTRODUCTION: There has been increased use of both induction of labor (IOL) and cesarean section for women with term pregnancies in many high-income countries, and a trend toward birth at earlier gestational ages. Existing evidence regarding the association between IOL and cesarean section for term pregnancies is mixed and conflicting, and little evidence is available on the differential effect at each week of gestation, stratified by parity. MATERIAL AND METHODS: To explore the association between IOL and primary cesarean section for singleton cephalic pregnancies at term, compared with two definitions of expectant management (first: at or beyond the week of gestation at birth following IOL; and secondary: only beyond the week of gestation at birth following IOL), we performed analyses of population-based historical cohort data on women who gave birth in one Australian state (Queensland), between July 1, 2012 and June 30, 2018. Women who gave birth before 37+0 or after 41+6 weeks of gestation, had stillbirths, no-labor, multiple births (twins or triplets), non-cephalic presentation at birth, a previous cesarean section, or missing data on included variables were excluded. Four sub-datasets were created for each week at birth (37-40). Unadjusted relative risk, adjusted relative risk using modified Poisson regression, and their 95% confidence intervals were calculated in each sub-dataset. Analyses were stratified by parity (nulliparas vs. parous women with a previous vaginal birth). Sensitivity analyses were conducted by limiting to women with low-risk pregnancies. RESULTS: A total of 239 094 women were included in the analysis, 36.7% of whom gave birth following IOL. The likelihood of primary cesarean section following IOL in a Queensland population-based cohort was significantly higher at 38 and 39 weeks, compared with expectant management up to 41+6 weeks, for both nulliparas and paras with singleton cephalic pregnancies, regardless of risk status of pregnancy and definition of expectant management. No significant difference was found for nulliparas at 37 and 40 weeks; and for paras at 40 weeks. CONCLUSIONS: Future studies are suggested to investigate further the association between IOL and other maternal and neonatal outcomes at each week of gestation in different maternal populations, before making any recommendation.

Identifying Longer-Term Health Events and Outcomes and Health Service Use of Low Birthweight CALD Infants in Australia.

INTRODUCTION: Approximately one-third of all births in Australia each year are by culturally and linguistically diverse (CALD) women. CALD women are at an increased risk of adverse pregnancy and birth outcomes including prematurity and low birthweight. Infants born weighing less than 2500 g are susceptible to increased risk of ill health and morbidities such as cognitive defects including cerebral palsy, and neuro-motor functioning. METHODS: An existing linked administrative dataset, Maternity 1000 was utilized for this study which has identified all children born in Queensland (QLD), Australia, between 1st July 2012 to 30th June 2018 from the QLD Perinatal Data Collection. This has then been linked to the QLD Hospital Admitted Patient Data Collection, QLD Hospital Non-Admitted Patient Data Collection, QLD Emergency Department Data Collection, and Medicare Benefits Schedule and Pharmaceutical Benefits Scheme Claims Records between 1 and 2012 to 30th June 2019. RESULTS: Culturally and linguistically diverse infants born with low birthweight had higher mean and standard deviation of all health events and outcomes; potentially preventable hospitalisations, hospital re-admissions, ED presentations without admissions, and development of chronic diseases compared to non-CALD infants born with low birthweight. DISCUSSION: Results from this study highlight the disparities in health service use and health events and outcomes associated with low birthweight infants, between both CALD and Australian born women. This study has responded to the knowledge gap of low birthweight on the Australian economy by identifying that there are significant inequalities in access to health services for CALD women in Australia, as well as increased health events and poor birth outcomes for these infants when compared to those of mothers born in Australia.

Assessing the fidelity of a behavioural intervention involving academic detailing in general practice: a sub-study of the 'Implementing work-related Mental health guidelines in general PRacticE' (IMPRovE) trial.

BACKGROUND: Assessing the fidelity of intervention components enables researchers to make informed judgements about the influence of those components on the observed outcome. The 'Implementing work-related Mental health guidelines in general PRacticE' (IMPRovE) trial is a hybrid III trial aiming to increase adherence to the 'Clinical Guidelines for the diagnosis and management of work-related mental health conditions in general practice'. IMPRovE is a multifaceted intervention, with one of the central components being academic detailing (AD). This study describes the fidelity to the protocol for the AD component of the IMPRovE intervention. METHOD: All AD sessions for the trial were audio-recorded and a sample of 22% were randomly selected for fidelity assessment. Fidelity was assessed using a tailored proforma based on the Modified Conceptual Framework for fidelity assessment, measuring duration, coverage, frequency and content. A descriptive analysis was used to quantify fidelity to the protocol and a content analysis was used to elucidate qualitative aspects of fidelity. RESULTS: A total of eight AD sessions were included in the fidelity assessment. The average fidelity score was 89.2%, ranging from 80 to 100% across the eight sessions. The sessions were on average 47 min long and addressed all of the ten chapters in the guideline. Of the guideline chapters, 9 were frequently discussed. The least frequently discussed chapter related to management of comorbid conditions. Most general practitioner (GP) participants used the AD sessions to discuss challenges with managing secondary mental conditions. In line with the protocol, opinion leaders who delivered the AD sessions largely offered evidence-based strategies aligning with the clinical guideline recommendations. CONCLUSIONS/IMPLICATIONS: The IMPRovE AD intervention component was delivered to high fidelity. The sessions adhered to the intended duration, coverage, frequency, and content allowing participating GPs to comprehend the implementation of the guideline in their own practice. This study also demonstrates that the Modified Conceptual Fidelity Framework with a mixed methods approach can support the assessment of implementation fidelity of a behavioural intervention in general practice. The findings enhance the trustworthiness of reported outcomes from IMPRovE and show that assessing fidelity is amenable for AD and should be incorporated in other studies using AD. TRIAL REGISTRATION: Australian New Zealand Clinical Trials Registry ACTRN 12620001163998, November 2020.