"Whatever journey you want to take, I'll support you through": a mixed methods evaluation of a peer worker program in the hospital emergency department.

BACKGROUND: People who are unhoused, use substances (drugs and/or alcohol), and who have mental health conditions experience barriers to care access and are frequently confronted with discrimination and stigma in health care settings. The role of Peer Workers in addressing these gaps in a hospital-based context is not well characterized. The aim of this evaluation was to 1) outline the role of Peer Workers in the care of a marginalized populations in the emergency department; 2) characterize the impact of Peer Workers on patient care, and 3) to describe how being employed as a Peer Worker impacts the Peer. METHODS: Through a concurrent mixed methods evaluation, we explore the role of Peer Workers in the care of marginalized populations in the emergency department at two urban hospitals in Toronto, Ontario Canada. We describe the demographic characteristics of patients (n = 555) and the type of supports provided to patients collected through a survey between February and June 2022. Semi-structured, in-depth interviews were completed with Peer Workers (n = 7). Interviews were thematically analyzed using a deductive approach, complemented by an inductive approach to allow new themes to emerge from the data. RESULTS: Support provided to patients primarily consisted of friendly conversations (91.4%), discharge planning (59.6%), tactics to help the patient navigate their emotions/mental wellbeing (57.8%) and sharing their lived experience (50.1%). In over one third (38.9%) of all patient interactions, Peer Workers shared new information about the patient with the health care team (e.g., obtaining patient identification). Five major themes emerged from our interviews with Peer Workers which include: (1) Establishing empathy and building trust between the patient and their care team through self-disclosure; (2) Facilitating a person-centered approach to patient care through trauma-informed listening and accessible language; (3) Support for patient preferences on harm reduction; (4) Peer worker role facilitating self-acceptance and self-defined recovery; and (5) Importance of supports and resources to help Peer Workers navigate the emotional intensity of the emergency department. CONCLUSIONS: The findings add to the literature on Peer Worker programs and how such interventions are designed to best meet the needs of marginalized populations.

Modelling population-level and targeted interventions of weight loss on chronic disease prevention in the Canadian population.

Obesity is a known risk factor for major chronic diseases. Prevention of chronic disease is a top global priority. The study aimed to model scenarios of population-level and targeted weight loss interventions on 10-year projected risk of chronic disease in Canada using a population-level risk prediction algorithm. The validated Chronic Disease Population Risk Tool (CDPoRT) forecasts 10-year risk of chronic disease in the adult population. We applied CDPoRT to the 2013/14 Canadian Community Health Survey to generate prospective chronic disease estimates for adults 20 years and older in Canada (n = 83,220). CDPoRT was used to model the following scenarios: British Columbia's (BC) and Quebec's (QC) provincial population-level weight reduction targets, a population-level intervention that could achieve weight loss, targeted weight loss interventions for overweight and obese groups, and the combination of a population-level and targeted weight loss intervention. We estimated chronic disease risk reductions and number of cases prevented in each scenario compared with the baseline. At baseline, we predicted an 18.4% risk and 4,151,929 new cases of chronic disease in Canada over the 10-year period. Provincial weight loss targets applied to the Canadian population estimated chronic disease reductions of 0.6% (BC) and 0.1% (QC). The population-level intervention estimated a greater reduction in risk (0.2%), compared to the targeted interventions (0.1%). The combined approach estimated a 0.3% reduction in chronic disease risk. Our modelling predicted that population-level approaches that achieve weight loss in combination with targeted weight loss interventions can substantially decrease the chronic disease burden in Canada.

Multimorbidity at time of death among persons with type 2 diabetes: a population-based study in Ontario, Canada.

OBJECTIVE: Individuals with Type 2 Diabetes are likely to experience multimorbidity and accumulate multiple chronic conditions over their life. We aimed to identify causes of death and chronic conditions at the time of death in a population-based cohort, and to analyze variations in the presence of diabetes at the time of death overall and across income and immigrant status. RESEARCH DESIGN AND METHODS: We conducted a retrospective cohort study of 2,199,801 adult deaths from 1992 to 2017 in Ontario, Canada. We calculated the proportion of decedents with chronic conditions at time of death and causes of death. The risk of diabetes at the time of death was modeled across sociodemographic variables with a log binomial regression adjusting for sex, age, immigrant status, area-level income. comorbiditiesand time. RESULTS: The leading causes of death in the cohort were cardiovascular and cancer. Decedents with diabetes had a higher prevalence of most chronic conditions than decedents without diabetes, including hypertension, osteo and other arthritis, chronic coronary syndrome, mood disorder, and congestive heart failure. The risk of diabetes at the time of death was 19% higher in immigrants (95%CI 1.18-1.20) and 15% higher in refugees (95%CI 1.12-1.18) compared to long-term residents, and 19% higher in the lowest income quintile (95%CI 1.18-1.20) relative to the highest income quintile, after adjusting for other covariates. CONCLUSIONS: Individuals with diabetes have a greater multimorbidity burden at the time of death, underscoring the importance of multiple chronic disease management among those living with diabetes and further considerations of the social determinants of health.

Predictive risk modelling of high resource users under different prescription drug coverage policies in Ontario and Manitoba, Canada.

INTRODUCTION: Studying high resource users (HRUs) across jurisdictions is a challenge due to variation in data availability and health services coverage. In Canada, coverage for pharmaceuticals varies across provinces under a mix of public and private plans, which has implications for ascertaining HRUs. We examined sociodemographic and behavioural predictors of HRUs in the presence of different prescription drug coverages in the provinces of Manitoba and Ontario. METHODS: Linked Canadian Community Health Surveys were used to create two cohorts of respondents from Ontario (n = 58,617, cycles 2005-2008) and Manitoba (n = 10,504, cycles 2007-2010). HRUs (top 5%) were identified by calculating health care utilization 5 years following interview date and computing all costs in the linked administrative databases, with three approaches used to include drug costs: (1) costs paid for by the provincial payer under age-based coverage; (2) costs paid for by the provincial payer under income-based coverage; (3) total costs regardless of the payer (publicly insured, privately insured, and out-of-pocket). Logistic regression estimated the association between sociodemographic, health, and behavioral predictors on HRU risk. RESULTS: The strength of the association between age (≥ 80 vs. <30) and becoming an HRU were attenuated with the inclusion of broader drug data (age based: OR 37.29, CI: 30.08-46.24; income based: OR 27.34, CI: 18.53-40.33; all drug payees: OR 29.08, CI: 19.64-43.08). With broader drug coverage, the association between heavy smokers vs. non-smokers on odds of becoming an HRU strengthened (age based: OR 1.58, CI: 1.32-1.90; income based: OR 2.97, CI: 2.18-4.05; all drug payees: OR 3.12, CI: 2.29-4.25). Across the different drug coverage policies, there was persistence in higher odds of becoming an HRU in low income households vs. high income households and in those with a reported chronic condition vs. no chronic conditions. CONCLUSIONS: The study illustrates that jurisdictional differences in how HRUs are ascertained based on drug coverage policies can influence the relative importance of some behavioural risk factors on HRU status, but most observed associations with health and sociodemographic risk factors were persistent, demonstrating that predictive risk modelling of HRUs can occur effectively across jurisdictions, even with some differences in public drug coverage policies.

Regional variation of premature mortality in Ontario, Canada: a spatial analysis.

BACKGROUND: Premature mortality is a meaningful indicator of both population health and health system performance, which varies by geography in Ontario. We used the Local Health Integration Network (LHIN) sub-regions to conduct a spatial analysis of premature mortality, adjusting for key population-level demographic and behavioural characteristics. METHODS: We used linked vital statistics data to identify 163,920 adult premature deaths (deaths between ages 18 and 74) registered in Ontario between 2011 and 2015. We compared premature mortality rates, population demographics, and prevalence of health-relevant behaviours across 76 LHIN sub-regions. We used Bayesian hierarchical spatial models to quantify the contribution of these population characteristics to geographic disparities in premature mortality. RESULTS: LHIN sub-region premature mortality rates ranged from 1.7 to 6.6 deaths per 1000 per year in males and 1.2 to 4.8 deaths per 1000 per year in females. Regions with higher premature mortality had fewer immigrants and higher prevalence of material deprivation, excess body weight, inadequate fruit and vegetable consumption, sedentary behaviour, and ever-smoked status. Adjusting for all variables eliminated close to 90% of geographic variation in premature mortality, but did not fully explain the spatial pattern of premature mortality in Ontario. CONCLUSIONS: We conducted the first spatial analysis of mortality in Ontario, revealing large geographic variations. We demonstrate that well-known risk factors explain most of the observed variation in premature mortality. The result emphasizes the importance of population health efforts to reduce the burden of well-known risk factors to reduce variation in premature mortality.

Predicting High Health Care Resource Utilization in a Single-payer Public Health Care System: Development and Validation of the High Resource User Population Risk Tool.

BACKGROUND: A large proportion of health care spending is incurred by a small proportion of the population. Population-based health planning tools that consider both the clinical and upstream determinants of high resource users (HRU) of the health system are lacking. OBJECTIVE: To develop and validate the High Resource User Population Risk Tool (HRUPoRT), a predictive model of adults that will become the top 5% of health care users over a 5-year period, based on self-reported clinical, sociodemographic, and health behavioral predictors in population survey data. RESEARCH DESIGN: The HRUPoRT model was developed in a prospective cohort design using the combined 2005 and 2007/2008 Canadian Community Health Surveys (CCHS) (N=58,617), and validated using the external 2009/2010 CCHS cohort (N=28,721). Health care utilization for each of the 5 years following CCHS interview date were determined by applying a person-centered costing algorithm to the linked health administrative databases. Discrimination and calibration of the model were assessed using c-statistic and Hosmer-Lemeshow (HL) χ statistic. RESULTS: The best prediction model for 5-year transition to HRU status included 12 predictors and had good discrimination (c-statistic=0.8213) and calibration (HL χ=18.71) in the development cohort. The model performed similarly in the validation cohort (c-statistic=0.8171; HL χ=19.95). The strongest predictors in the HRUPoRT model were age, perceived general health, and body mass index. CONCLUSIONS: HRUPoRT can accurately project the proportion of individuals in the population that will become a HRU over 5 years. HRUPoRT can be applied to inform health resource planning and prevention strategies at the community level.

Characterising violent deaths of undetermined intent: a population-based study, 1999-2012.

OBJECTIVES: Violent deaths classified as undetermined intent (UD) are sometimes included in suicide counts. This study investigated age and sex differences, along with socioeconomic gradients in UD and suicide deaths in the province of Ontario between 1999 and 2012. METHODS: We used data from the Institute for Clinical Evaluative Sciences, which has linked vital statistics from the Office of the Registrar General Deaths register with Census data between 1999 and 2012. Socioeconomic status was operationalised through the four dimensions of the Ontario Marginalization Index. We computed age-specific and annual age-standardised mortality rates, and risk ratios to calculate risk gradients according to each of the four dimensions of marginalization. RESULTS: Rates of UD-classified deaths were highest for men aged 45-64 years residing in the most materially deprived (7.9 per 100 000 population (95% CI 6.8 to 9.0)) and residentially unstable (8.1 (95% CI 7.1 to 9.1)) neighbourhoods. Similarly, suicide rates were highest among these same groups of men aged 45-64 living in the most materially deprived (28.2 (95% CI 26.1 to 30.3)) and residentially unstable (30.7 (95% CI 28.7 to 32.6)) neighbourhoods. Relative to methods of death, poisoning was the most frequently used method in UD cases (64%), while it represented the second most common method (27%) among suicides after hanging (40%). DISCUSSION: The similarities observed between both causes of death suggest that at least a proportion of UD deaths may be misclassified suicide cases. However, the discrepancies identified in this analysis seem to indicate that not all UD deaths are misclassified suicides.

Putting a Population Health Lens to Multimorbidity in Ontario.

Almost all Ontarians die with multimorbidity, and most accumulate more than five conditions over their lifetime. Our health system is still largely focused on specialties and treating one disease at a time - an approach that is incompatible with the healthcare needs of patients with multiple and often complex chronic conditions. This burden requires a health system that recognizes that patients will more likely live and die with multiple chronic conditions than not (i.e., multimorbidity management versus specialized care). There are important and meaningful differences in the types and numbers of conditions that patients die with. In particular, increases in the most preventable conditions are greater among the most deprived members of our society. To address the worrying trends seen here, chronic disease prevention - not only management - must be a priority, with a strong focus on health equity. Chronic disease prevention and a strong focus on equity are signatures of a population health approach. This work echoes calls for a stronger emphasis on population health in the health system.

A population-based study of homicide deaths in Ontario, Canada using linked death records.

BACKGROUND: Homicide - a lethal expression of violence - has garnered little attention from public health researchers and health policy makers, despite the fact that homicides are a cause of preventable and premature death. Identifying populations at risk and the upstream determinants of homicide are important for addressing inequalities that hinder population health. This population-based study investigates the public health significance of homicides in Ontario, Canada, over the period of 1999-2012. We quantified the relative burden of homicides by comparing the socioeconomic gradient in homicides with the leading causes of death, cardiovascular disease (CVD) and neoplasm, and estimated the potential years of life lost (PYLL) due to homicide. METHODS: We linked vital statistics from the Office of the Registrar General Deaths register (ORG-D) with Census and administrative data for all Ontario residents. We extracted all homicide, neoplasm, and cardiovascular deaths from 1999 to 2012, using International Classification of Diseases codes. For socioeconomic status (SES), we used two dimensions of the Ontario Marginalization Index (ON-Marg): material deprivation and residential instability. Trends were summarized across deprivation indices using age-specific rates, rate ratios, and PYLL. RESULTS: Young males, 15-29 years old, were the main victims of homicide with a rate of 3.85 [IC 95%: 3.56; 4.13] per 100,000 population and experienced an upward trend over the study period. The socioeconomic neighbourhood gradient was substantial and higher than the gradient for both cardiovascular and neoplasms. Finally, the PYLL due to homicide were 63,512 and 24,066 years for males and females, respectively. CONCLUSIONS: Homicides are an important cause of death among young males, and populations living in disadvantaged neighbourhoods. Our findings raise concerns about the burden of homicides in the Canadian population and the importance of addressing social determinants to address these premature deaths.

Using the Nine Common Themes of Good Practice checklist as a tool for evaluating the research priority setting process of a provincial research and program evaluation program.

BACKGROUND: Given the context-specific nature of health research prioritization and the obligation to effectively allocate resources to initiatives that will achieve the greatest impact, evaluation of priority setting processes can refine and strengthen such exercises and their outcomes. However, guidance is needed on evaluation tools that can be applied to research priority setting. This paper describes the adaption and application of a conceptual framework to evaluate a research priority setting exercise operating within the public health sector in Ontario, Canada. METHODS: The Nine Common Themes of Good Practice checklist, described by Viergever et al. (Health Res Policy Syst 8:36, 2010) was used as the conceptual framework to evaluate the research priority setting process developed for the Locally Driven Collaborative Projects (LDCP) program in Ontario, Canada. Multiple data sources were used to inform the evaluation, including a review of selected priority setting approaches, surveys with priority setting participants, document review, and consultation with the program advisory committee. RESULTS: The evaluation assisted in identifying improvements to six elements of the LDCP priority setting process. The modifications were aimed at improving inclusiveness, information gathering practices, planning for project implementation, and evaluation. In addition, the findings identified that the timing of priority setting activities and level of control over the process were key factors that influenced the ability to effectively implement changes. CONCLUSIONS: The findings demonstrate the novel adaptation and application of the 'Nine Common Themes of Good Practice checklist' as a tool for evaluating a research priority setting exercise. The tool can guide the development of evaluation questions and enables the assessment of key constructs related to the design and delivery of a research priority setting process.

External validation and application of the Diabetes Population Risk Tool (DPoRT) for prediction of type 2 diabetes onset in the US population.

INTRODUCTION: Characterizing diabetes risk in the population is important for population health assessment and diabetes prevention planning. We aimed to externally validate an existing 10-year population risk model for type 2 diabetes in the USA and model the population benefit of diabetes prevention approaches using population survey data. RESEARCH DESIGN AND METHODS: The Diabetes Population Risk Tool (DPoRT), originally derived and validated in Canada, was applied to an external validation cohort of 23 477 adults from the 2009 National Health Interview Survey (NHIS). We assessed predictive performance for discrimination (C-statistic) and calibration plots against observed incident diabetes cases identified from the NHIS 2009-2018 cycles. We applied DPoRT to the 2018 NHIS cohort (n=21 187) to generate 10-year risk prediction estimates and characterize the preventive benefit of three diabetes prevention scenarios: (1) community-wide strategy; (2) high-risk strategy and (3) combined approach. RESULTS: DPoRT demonstrated good discrimination (C-statistic=0.778 (males); 0.787 (females)) and good calibration across the range of risk. We predicted a baseline risk of 10.2% and 21 076 000 new cases of diabetes in the USA from 2018 to 2028. The community-wide strategy and high-risk strategy estimated diabetes risk reductions of 0.2% and 0.3%, respectively. The combined approach estimated a 0.4% risk reduction and 843 000 diabetes cases averted in 10 years. CONCLUSIONS: DPoRT has transportability for predicting population-level diabetes risk in the USA using routinely collected survey data. We demonstrate the model's applicability for population health assessment and diabetes prevention planning. Our modeling predicted that the combination of community-wide and targeted prevention approaches for those at highest risk are needed to reduce diabetes burden in the USA.

Modeling chronic disease risk across equity factors using a population-based prediction model: the Chronic Disease Population Risk Tool (CDPoRT).

BACKGROUND: Predicting chronic disease incidence at a population level can help inform overall future chronic disease burden and opportunities for prevention. This study aimed to estimate the future burden of chronic disease in Ontario, Canada, using a population-level risk prediction algorithm and model interventions for equity-deserving groups who experience barriers to services and resources due to disadvantages and discrimination. METHODS: The validated Chronic Disease Population Risk Tool (CDPoRT) estimates the 10-year risk and incidence of major chronic diseases. CDPoRT was applied to data from the 2017/2018 Canadian Community Health Survey to predict baseline 10-year chronic disease estimates to 2027/2028 in the adult population of Ontario, Canada, and among equity-deserving groups. CDPoRT was used to model prevention scenarios of 2% and 5% risk reductions over 10 years targeting high-risk equity-deserving groups. RESULTS: Baseline chronic disease risk was highest among those with less than secondary school education (37.5%), severe food insecurity (19.5%), low income (21.2%) and extreme workplace stress (15.0%). CDPoRT predicted 1.42 million new chronic disease cases in Ontario from 2017/2018 to 2027/2028. Reducing chronic disease risk by 5% prevented 1500 cases among those with less than secondary school education, prevented 14 900 cases among those with low household income and prevented 2800 cases among food-insecure populations. Large reductions of 57 100 cases were found by applying a 5% risk reduction in individuals with quite a bit workplace stress. CONCLUSION: Considerable reduction in chronic disease cases was predicted across equity-defined scenarios, suggesting the need for prevention strategies that consider upstream determinants affecting chronic disease risk.

A study protocol for a predictive model to assess population-based avoidable hospitalization risk: Avoidable Hospitalization Population Risk Prediction Tool (AvHPoRT).

INTRODUCTION: Avoidable hospitalizations are considered preventable given effective and timely primary care management and are an important indicator of health system performance. The ability to predict avoidable hospitalizations at the population level represents a significant advantage for health system decision-makers that could facilitate proactive intervention for ambulatory care-sensitive conditions (ACSCs). The aim of this study is to develop and validate the Avoidable Hospitalization Population Risk Tool (AvHPoRT) that will predict the 5-year risk of first avoidable hospitalization for seven ACSCs using self-reported, routinely collected population health survey data. METHODS AND ANALYSIS: The derivation cohort will consist of respondents to the first 3 cycles (2000/01, 2003/04, 2005/06) of the Canadian Community Health Survey (CCHS) who are 18-74 years of age at survey administration and a hold-out data set will be used for external validation. Outcome information on avoidable hospitalizations for 5 years following the CCHS interview will be assessed through data linkage to the Discharge Abstract Database (1999/2000-2017/2018) for an estimated sample size of 394,600. Candidate predictor variables will include demographic characteristics, socioeconomic status, self-perceived health measures, health behaviors, chronic conditions, and area-based measures. Sex-specific algorithms will be developed using Weibull accelerated failure time survival models. The model will be validated both using split set cross-validation and external temporal validation split using cycles 2000-2006 compared to 2007-2012. We will assess measures of overall predictive performance (Nagelkerke R2), calibration (calibration plots), and discrimination (Harrell's concordance statistic). Development of the model will be informed by the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) statement. ETHICS AND DISSEMINATION: This study was approved by the University of Toronto Research Ethics Board. The predictive algorithm and findings from this work will be disseminated at scientific meetings and in peer-reviewed publications.

Variations in all-cause mortality, premature mortality and cause-specific mortality among persons with diabetes in Ontario, Canada.

INTRODUCTION: Patients with diabetes have a higher risk of mortality compared with the general population. Large population-based studies that quantify variations in mortality risk for patients with diabetes among subgroups in the population are lacking. This study aimed to examine the sociodemographic differences in the risk of all-cause mortality, premature mortality, and cause-specific mortality in persons diagnosed with diabetes. RESEARCH DESIGN AND METHODS: We conducted a population-based cohort study of 1 741 098 adults diagnosed with diabetes between 1994 and 2017 in Ontario, Canada using linked population files, Canadian census, health administrative and death registry databases. We analyzed the association between sociodemographics and other covariates on all-cause mortality and premature mortality using Cox proportional hazards models. A competing risk analysis using Fine-Gray subdistribution hazards models was used to analyze cardiovascular and circular mortality, cancer mortality, respiratory mortality, and mortality from external causes of injury and poisoning. RESULTS: After full adjustment, individuals with diabetes who lived in the lowest income neighborhoods had a 26% (HR 1.26, 95% CI 1.25 to 1.27) increased hazard of all-cause mortality and 44% (HR 1.44, 95% CI 1.42 to 1.46) increased risk of premature mortality, compared with individuals with diabetes living in the highest income neighborhoods. In fully adjusted models, immigrants with diabetes had reduced risk of all-cause mortality (HR 0.46, 95% CI 0.46 to 0.47) and premature mortality (HR 0.40, 95% CI 0.40 to 0.41), compared with long-term residents with diabetes. Similar HRs associated with income and immigrant status were observed for cause-specific mortality, except for cancer mortality, where we observed attenuation in the income gradient among persons with diabetes. CONCLUSIONS: The observed mortality variations suggest a need to address inequality gaps in diabetes care for persons with diabetes living in the lowest income areas.

Experiences and Reported Outcomes of Patients and Caregivers Enrolled in an Integrated Care Program for Thoracic Surgery: A Qualitative Evaluation.

Background: Health care delivery is often poorly coordinated and fragmented. Integrated care (IC) programs represent one solution to improving continuity of care. The aim of this study was to understand experiences and reported outcomes of patients and caregivers in an IC Program that coordinates hospital and home care for thoracic surgery. Methods: A process evaluation was undertaken using qualitative methods. We conducted semi-structured interviews with 10 patients and 8 caregivers who received IC for thoracic surgery and were discharged between June 2019 and April 2020. A phenomenological approach was used to understand and characterize patient and caregiver experiences. Thematic analysis began with a deductive approach complemented by an inductive approach. Results: Four major themes evolved from patient and caregiver interviews, including 1) coordination and timeliness of patient care facilitated by an IC lead; 2) the provision of person-centred care and relational continuity fostered feelings of partnership with patients and caregivers; 3) clear communication and one shared digital record increased informational continuity; and 4) impacts of IC on patient and caregiver outcomes. Conclusions: Patients and caregivers generally reported this IC Program met their health care needs, which may help inform how future IC programs are designed.

Predicting hospitalisations related to ambulatory care sensitive conditions with machine learning for population health planning: derivation and validation cohort study.

OBJECTIVE: To predict older adults' risk of avoidable hospitalisation related to ambulatory care sensitive conditions (ACSC) using machine learning applied to administrative health data of Ontario, Canada. DESIGN, SETTING AND PARTICIPANTS: A retrospective cohort study was conducted on a large cohort of all residents covered under a single-payer system in Ontario, Canada over the period of 10 years (2008-2017). The study included 1.85 million Ontario residents between 65 and 74 years old at any time throughout the study period. DATA SOURCES: Administrative health data from Ontario, Canada obtained from the (ICES formely known as the Institute for Clinical Evaluative Sciences Data Repository. MAIN OUTCOME MEASURES: Risk of hospitalisations due to ACSCs 1 year after the observation period. RESULTS: The study used a total of 1 854 116 patients, split into train, validation and test sets. The ACSC incidence rates among the data points were 1.1% for all sets. The final XGBoost model achieved an area under the receiver operating curve of 80.5% and an area under precision-recall curve of 0.093 on the test set, and the predictions were well calibrated, including in key subgroups. When ranking the model predictions, those at the top 5% of risk as predicted by the model captured 37.4% of those presented with an ACSC-related hospitalisation. A variety of features such as the previous number of ambulatory care visits, presence of ACSC-related hospitalisations during the observation window, age, rural residence and prescription of certain medications were contributors to the prediction. Our model was also able to capture the geospatial heterogeneity of ACSC risk in Ontario, and especially the elevated risk in rural and marginalised regions. CONCLUSIONS: This study aimed to predict the 1-year risk of hospitalisation from ambulatory-care sensitive conditions in seniors aged 65-74 years old with a single, large-scale machine learning model. The model shows the potential to inform population health planning and interventions to reduce the burden of ACSC-related hospitalisations.

Forecasting Diabetes Cases Prevented and Cost Savings Associated with Population Increases of Walking in the Greater Toronto and Hamilton Area, Canada.

Promoting adequate levels of physical activity in the population is important for diabetes prevention. However, the scale needed to achieve tangible population benefits is unclear. We aimed to estimate the public health impact of increases in walking as a means of diabetes prevention and health care cost savings attributable to diabetes. We applied the validated Diabetes Population Risk Tool (DPoRT) to the 2015/16 Canadian Community Health Survey for adults aged 18-64, living in the Greater Toronto and Hamilton area, Ontario, Canada. DPoRT was used to generate three population-level scenarios involving increases in walking among individuals with low physical activity levels, low daily step counts and high dependency on non-active forms of travel, compared to a baseline scenario (no change in walking rates). We estimated number of diabetes cases prevented and health care costs saved in each scenario compared with the baseline. Each of the three scenarios predicted a considerable reduction in diabetes and related health care cost savings. In order of impact, the largest population benefits were predicted from targeting populations with low physical activity levels, low daily step counts, and non active transport use. Population increases of walking by 25 min each week was predicted to prevent up to 10.4 thousand diabetes cases and generate CAD 74.4 million in health care cost savings in 10 years. Diabetes reductions and cost savings were projected to be higher if increases of 150 min of walking per week could be achieved at the population-level (up to 54.3 thousand diabetes cases prevented and CAD 386.9 million in health care cost savings). Policy, programming, and community designs that achieve modest increases in population walking could translate to meaningful reductions in the diabetes burden and cost savings to the health care system.

Estimating Population Benefits of Prevention Approaches Using a Risk Tool: High Resource Users in Ontario, Canada.

BACKGROUND: Healthcare spending is concentrated, with a minority of the population accounting for the majority of healthcare costs. METHODS: The authors modelled the impact of high resource user (HRU) prevention strategies within five years using the validated High Resource User Population Risk Tool. RESULTS: The authors estimated 758,000 new HRUs in Ontario from 2013-2014 to 2018-2019, resulting in $16.20 billion in healthcare costs (Canadian dollars 2016). The prevention approach that had the largest reduction in HRUs was targeting health-risk behaviours. CONCLUSIONS: This study demonstrates the use of a policy tool by decision makers to support prevention approaches that consider the impact on HRUs and estimated healthcare costs.

Characterising mental health and addictions and assault-related health care use in the year prior to death: a population-based linked cohort study of homicide victims.

INTRODUCTION: Homicide is an important cause of death for older youth and adult Canadians; however, little is known about health care use prior to death among this population. OBJECTIVES: To characterise health care use for mental health and addictions (MHA) and serious assault (herein referred to assault) one year prior to death among individuals who died by homicide in Ontario, Canada using linked mortality and health care utilisation data. METHODS: We report rates of health care use for MHA and assault in the year prior to death among all individuals 16 years and older in Ontario, Canada, who died by homicide from April 2003 to December 2012 (N = 1,541). Health care use for MHA included inpatient stays, emergency department (ED) visits and outpatient visits, whereas health care use for assault included only hospital-based care (ED visits and inpatient stays). Sociodemographic characteristics and health care utilisation were examined across homicide deaths, stratified by sex. RESULTS: Overall, 28.5% and 5.9% of homicide victims sought MHA and assault care in the year prior to death, respectively. A greater proportion of females accessed care for MHA, whereas a greater proportion of males accessed assault-related health care. Males were more likely to be hospitalised following an ED visit for a MHA or assault related reason, in comparison to females. The most common reason for a MHA hospital visit was for substance-related disorders. We found an increase over time for hospital-based visits for assault prior to death, a trend that was not observed for MHA-related visits. CONCLUSIONS: A large proportion of homicide victims interacted with the health care system for MHA or assault in the year prior to death. An increase in hospital-based visits for assault-related reasons prior to death was observed. These trends may offer insight into avenues for support and prevention for victims of homicide.

Development and Validation of a Machine Learning Model Using Administrative Health Data to Predict Onset of Type 2 Diabetes.

Importance: Systems-level barriers to diabetes care could be improved with population health planning tools that accurately discriminate between high- and low-risk groups to guide investments and targeted interventions. Objective: To develop and validate a population-level machine learning model for predicting type 2 diabetes 5 years before diabetes onset using administrative health data. Design, Setting, and Participants: This decision analytical model study used linked administrative health data from the diverse, single-payer health system in Ontario, Canada, between January 1, 2006, and December 31, 2016. A gradient boosting decision tree model was trained on data from 1 657 395 patients, validated on 243 442 patients, and tested on 236 506 patients. Costs associated with each patient were estimated using a validated costing algorithm. Data were analyzed from January 1, 2006, to December 31, 2016. Exposures: A random sample of 2 137 343 residents of Ontario without type 2 diabetes was obtained at study start time. More than 300 features from data sets capturing demographic information, laboratory measurements, drug benefits, health care system interactions, social determinants of health, and ambulatory care and hospitalization records were compiled over 2-year patient medical histories to generate quarterly predictions. Main Outcomes and Measures: Discrimination was assessed using the area under the receiver operating characteristic curve statistic, and calibration was assessed visually using calibration plots. Feature contribution was assessed with Shapley values. Costs were estimated in 2020 US dollars. Results: This study trained a gradient boosting decision tree model on data from 1 657 395 patients (12 900 257 instances; 6 666 662 women [51.7%]). The developed model achieved a test area under the curve of 80.26 (range, 80.21-80.29), demonstrated good calibration, and was robust to sex, immigration status, area-level marginalization with regard to material deprivation and race/ethnicity, and low contact with the health care system. The top 5% of patients predicted as high risk by the model represented 26% of the total annual diabetes cost in Ontario. Conclusions and Relevance: In this decision analytical model study, a machine learning model approach accurately predicted the incidence of diabetes in the population using routinely collected health administrative data. These results suggest that the model could be used to inform decision-making for population health planning and diabetes prevention.