Search

We introduce a novel method to estimate the causal effects of an intervention over multiple treated units by combining the techniques of probabilistic forecasting with global forecasting methods using deep learning (DL) models. Considering the counterfactual and synthetic approach for policy evaluation, we recast the causal effect estimation problem as a counterfactual prediction outcome of the treated units in the absence of the treatment. Nevertheless, in contrast to estimating only the counterfactual time series outcome, our work differs from conventional methods by proposing to estimate the counterfactual time series probability distribution based on the past preintervention set of treated and untreated time series. We rely on time series properties and forecasting methods, with shared parameters, applied to stacked univariate time series for causal identification. This article presents DeepProbCP, a framework for producing accurate quantile probabilistic forecasts for the counterfactual outcome, based on training a global autoregressive recurrent neural network model with conditional quantile functions on a large set of related time series. The output of the proposed method is the counterfactual outcome as the spline-based representation of the counterfactual distribution. We demonstrate how this probabilistic methodology added to the global DL technique to forecast the counterfactual trend and distribution outcomes overcomes many challenges faced by the baseline approaches to the policy evaluation problem. Oftentimes, some target interventions affect only the tails or the variance of the treated units' distribution rather than the mean or median, which is usual for skewed or heavy-tailed distributions. Under this scenario, the classical causal effect models based on counterfactual predictions are not capable of accurately capturing or even seeing policy effects. By means of empirical evaluations of synthetic and real-world datasets, we show that our framework delivers more accurate forecasts than the state-of-the-art models, depicting, in which quantiles, the intervention most affected the treated units, unlike the conventional counterfactual inference methods based on nonprobabilistic approaches.

Gamma-hydroxybutyrate (GHB) use and attributable harms have been increasing in Australia, however changes over time, including the impact of COVID-19 lockdowns and restrictions on harms requiring an ambulance attendance, are unknown. This study utilised a novel population-based surveillance system to identify the types of GHB-related harms between January 2018 and 31 December 2021 in Victoria, Australia.A cross-sectional, retrospective analysis of all GHB-related ambulance attendances between January 2018 and 31 December 2021 in Victoria, Australia was undertaken. Paramedic clinical notes and Glasgow Coma Scale scores were used to assess conscious state. Event codes were classified using dispatch information available in the database. Crude rates (per 100,000 population) and descriptive analyses were calculated for metropolitan and regional settings. Adjusted Odds ratios (aOR) with 95% confidence intervals [95% CI] were used to assess the relationship between GCS severity and polysubstance combinations with GHB.There were 6,836 ambulance attendances for GHB recorded during the study period. A statistically significant increase in GHB-related attendance numbers was observed State-wide in 2019 (n = 1,402, p0.001) and 2020 (n = 2,622, p0.001), when comparing year on year attendances. While both numbers and rates (per 100,000 population) of GHB-related attendances were significantly lower in regional areas, significant increases were evident in both metropolitan and regional areas in 2019 and 2020 (both p0.001). Attendances involving GHB and alcohol had higher odds of a severe GCS score (aOR:1.25; 95%: 1.04-1.49; p0.019). A high proportion of GHB-attendances involved harms of significant concern including: overdose (56%) and a loss of, or altered state of consciousness (45%).We observed increases in GHB-related ambulance attendances over time in both metropolitan and regional areas, placing a significant burden on ambulance services. Our study demonstrates the value of using ambulance surveillance to obtain representative data on acute GHB-related harms.

Alcohol harms are often determined using a proxy measure based on temporal patterns during the week when harms are most likely to occur. This study utilised coded Australian ambulance data from the Victorian arm of the National Ambulance Surveillance System (NASS) to investigate temporal patterns across the week for alcohol-related ambulance attendances in 2019. These patterns were examined by season, regionality, gender, and age group. We found clear temporal peaks: from Friday 6:00 p.m. to Saturday 3:59 a.m. for both alcohol-involved and alcohol-intoxication-related attendance, from Saturday 6:00 p.m. to Sunday 4:59 a.m. for alcohol-involved attendances, and from Saturday 5:00 p.m. to Sunday 4:49 a.m. for alcohol-intoxication-related attendances. However, these temporal trends varied across age groups. Additionally, hours during Thursday and Sunday evenings also demonstrated peaks in attendances. There were no substantive differences between genders. Younger age groups (18–24 and 25–29 years) had a peak of alcohol-related attendances from 7:00 p.m. to 7:59 a.m. on Friday and Saturday nights, whereas the peak in attendances for 50–59 and 60+ years was from 5:00 p.m. to 2:59 a.m. on Friday and Saturday nights. These findings further the understanding of the impacts of alcohol during different times throughout the week, which can guide targeted policy responses regarding alcohol use and health service capacity planning.

IntroductionThe Banned Drinker Register (BDR) was reintroduced in the Northern Territory (NT) in September 2017. The BDR is a supply reduction measure and involves placing people who consume alcohol at harmful levels on a register prohibiting the purchase, possession and consumption of alcohol. The current study aims to evaluate the impacts of the reintroduction of the BDR, in the context of other major alcohol policy initiatives introduced across the NT such as Police Auxiliary Liquor Inspectors and a minimum unit price for alcohol of US$1.30 per standard drink.Methods and analysesThe Learning from Alcohol (policy) Reforms in the Northern Territory project will use a mixed-methods approach and contain four major components: epidemiological analysis of trends over time (outcomes include health, justice and social welfare data); individual-level data linkage including those on the BDR (outcomes include health and justice data); qualitative interviews with key stakeholders in the NT (n≥50); and qualitative interviews among people who are, or were previously, on the BDR, as well as the families and communities connected to those on the BDR (n=150). The impacts of the BDR on epidemiological data will be examined using time series analysis. Linked data will use generalised mixed models to analyse the relationship between outcomes and exposures, utilising appropriate distributions. Qualitative data will be analysed using thematic analysis.Ethics and disseminationEthics approvals have been obtained from NT Department of Health and Menzies School of Health Research Human Research Ethics Committee (HREC), Central Australia HREC and Deakin University HREC. In addition to peer-reviewed publications, we will report our findings to key organisational, policy, government and community stakeholders via conferences, briefings and lay summaries.

Substance use is a risk factor for being both a perpetrator and a victim of violence. The aim of this systematic review was to report the prevalence of acute pre-injury substance use in patients with violence-related injuries. Systematic searches were used to identify observational studies that included patients aged ≥15 years presenting to hospital after violence-related injuries and used objective toxicology measures to report prevalence of acute pre-injury substance use. Studies were grouped based on injury cause (any violence-related, assault, firearm, and other penetrating injuries including stab and incised wounds) and substance type (any substance, alcohol only, drugs other than alcohol only), and they were summarized using narrative synthesis and meta-analyses. This review included 28 studies. Alcohol was detected in 13%–66% of any violence-related injuries (five studies), 4%–71% of assaults (13 studies), 21%–45% of firearm injuries (six studies; pooled estimate = 41%, 95% CI: 40%–42%, n = 9,190), and 9%–66% of other penetrating injuries (nine studies; pooled estimate = 60%, 95% CI: 56%–64%, n = 6,950). Drugs other than alcohol were detected in 37% of any violence-related injuries (one study), 39% of firearm injuries (one study), 7%–49% of assaults (five studies), and 5%–66% of penetrating injuries (three studies). The prevalence of any substance varied across injury categories: any violence-related injuries = 76%–77% (three studies), assaults = 40%–73% (six studies), firearms = n/a, other penetrating injuries = 26%–45% (four studies; pooled estimate = 30%, 95% CI: 24%–37%, n = 319).Overall, substance use was frequently detected in patients presenting to hospital for violence-related injuries. Quantification of substance use in violence-related injuries provides a benchmark for harm reduction and injury prevention strategies.

Objective: To identify processes to engage stakeholders in healthcare Simulation Modeling (SM), and the impacts of this engagement on model design, model implementation, and stakeholder participants. To investigate how engagement process may lead to specific impacts.Data Sources: English-language articles on health SM engaging stakeholders in the MEDLINE, EMBASE, Scopus, Web of Science and Business Source Complete databases published from inception to February 2020.Study Design: A systematic review of the literature based on a priori protocol and reported according to PRISMA guidelines.Extraction Methods: Eligible articles were SM studies with a health outcome which engaged stakeholders in model design. Data were extracted using a data extraction form adapted to be specific for stakeholder engagement in SM studies. Data were analyzed using summary statistics, deductive and inductive content analysis, and narrative synthesis.Principal Findings: Thirty-two articles met inclusion criteria. Processes used to engage stakeholders in healthcare SM are heterogenous and often based on intuition rather than clear methodological frameworks. These processes most commonly involve stakeholders across multiple SM stages via discussion/dialogue, interviews, workshops and meetings. Key reported impacts of stakeholder engagement included improved model quality/accuracy, implementation, and stakeholder decision-making. However, for all but four studies, these reports represented author perceptions rather than formal evaluations incorporating stakeholder perspectives. Possible process enablers of impact included the use of models as “boundary objects” and structured facilitation via storytelling to promote effective communication and mutual understanding between stakeholders and modelers.Conclusions: There is a large gap in the current literature of formal evaluation of SM stakeholder engagement, and a lack of consensus about the processes required for effective SM stakeholder engagement. The adoption and clear reporting of structured engagement and process evaluation methodologies/frameworks are required to advance the field and produce evidence of impact.