Search

Abstract Outpatient care patterns have changed markedly during the COVID-19 pandemic. In this population-based retrospective cohort study, we compared the frequency of outpatient care (whether in-person or virtual) and continuity of care for all community-dwelling adults in Alberta between March 1, 2019 and February 29, 2020 (pre-pandemic) versus March 1, 2020 to February 28, 2021 (pandemic). We calculated provider continuity using Breslau’s Usual Provider Continuity (UPC) for patients with at least 2 outpatient encounters. In 2019–20, 594,350 (98.4%) of 603,877 community-dwelling adults with ambulatory care sensitive conditions (ACSC) had $$\ge$$ ≥ 1 outpatient visit (median 8 visits, mean UPC score 0.61, SD 0.23), compared to 566,569 (98.6%) of 574,613 (median 8 visits, mean UPC score 0.67, SD 0.23) during the first year of the pandemic. Similar patterns were seen for adults without ACSC: 2,207,710 (93.9%) of 2,350,147 had $$\ge$$ ≥ 1 outpatient visit (median 3 visits, mean UPC score 0.61, SD 0.24) pre-pandemic compared to 2,113,239 (93.5%, median 4 visits, mean UPC 0.67, SD 0.24) in the first year of the pandemic. Thus, the COVID-19 pandemic did not impact frequency of follow-up while continuity of care improved both for patients with or without ACSC in Alberta, Canada.

Background: Pulmonary hypertension is common among patients with heart failure (HF). Right ventricular systolic pressure (RVSP) is frequently used to assess its presence and severity. Although RVSP has been associated with adverse outcomes, the importance of serial measurements has not been studied. We evaluated associations between serial RVSP measurements and cardiovascular events in patients with HF. Methods: Patients with HF and 2 echocardiograms performed ≥ 6 months apart were included. RVSP was categorized, using the second echocardiogram, as follows: normal (< 40 mm Hg); severely elevated (≥ 60 mm Hg); moderately elevated (50-59 mm Hg); or mildly elevated (40-49 mm Hg). Patients also were classified according to change in RVSP categories between echocardiograms. The primary outcome was time to HF hospitalization (HFH) or all-cause mortality (ACM) after the second echocardiogram. Results: In total, 4319 patients were included (median age: 78 years; 52.1% female). During a median follow-up period of 19.4 months, HFH/ACM occurred in 2714 patients (62.8%). In multivariable analysis, baseline RSVP that was mildly elevated (1069 patients, hazard ratio [HR] 1.31, 95% confidence interval [CI] 1.12-1.54), moderately elevated (797 patients, HR 1.54, 95% CI 1.30-1.82), or severely elevated (837 patients, HR 1.92, 95% CI 1.60-2.31) was independently associated with HFH/ACM. Additionally, improving RVSP was associated with increased HFH/ACM in both categorical (HR 1.16, 95% CI 1.01-1.33) and continuous analyses. Conclusions: RVSP measurements identify patients at increased risk who may require more-aggressive monitoring and medical therapy. Our study raises the hypothesis that, in addition to the absolute value of RVSP, improving RVSP category may identify higher-risk patients, but further study is needed to elucidate the underlying reasons. Résumé: Contexte: L’hypertension pulmonaire est fréquente chez les patients atteints d’insuffisance cardiaque. La pression systolique ventriculaire droite (PSVD) est souvent utilisée pour évaluer la présence et la gravité de l’hypertension pulmonaire. Bien que la PSVD ait été associée à des complications, l’importance de mesures répétitives n’a pas été étudiée. Nous avons évalué les liens entre des mesures répétitives de la PSVD et des événements cardiovasculaires chez des patients atteints d’insuffisance cardiaque. Méthodologie: Ont été inclus des patients atteints d’insuffisance cardiaque pour lesquels on disposait de deux échocardiogrammes réalisés dans un intervalle ≥ 6 mois. La PSVD a été catégorisée comme suit, au moyen du deuxième échocardiogramme : normale (

Abstract Many health authorities differentiate hospitalizations in patients infected with SARS-CoV-2 as being “for COVID-19” (due to direct manifestations of SARS-CoV-2 infection) versus being an “incidental” finding in someone admitted for an unrelated condition. We conducted a retrospective cohort study of all SARS-CoV-2 infected patients hospitalized via 47 Canadian emergency departments, March 2020-July 2022 to determine whether hospitalizations with “incidental” SARS-CoV-2 infection are less of a burden to patients and the healthcare system. Using a priori standardized definitions applied to hospital discharge diagnoses in 14,290 patients, we characterized COVID-19 as (i) the “Direct” cause for the hospitalization (70%), (ii) a potential “Contributing” factor for the hospitalization (4%), or (iii) an “Incidental” finding that did not influence the need for admission (26%). The proportion of incidental SARS-CoV-2 infections rose from 10% in Wave 1 to 41% during the Omicron wave. Patients with COVID-19 as the direct cause of hospitalization exhibited significantly longer LOS (mean 13.8 versus 12.1 days), were more likely to require critical care (22% versus 11%), receive COVID-19-specific therapies (55% versus 19%), and die (17% versus 9%) compared to patients with Incidental SARS-CoV-2 infections. However, patients hospitalized with incidental SARS-CoV-2 infection still exhibited substantial morbidity/mortality and hospital resource use.

Abstract The feasibility and value of linking electrocardiogram (ECG) data to longitudinal population-level administrative health data to facilitate the development of a learning healthcare system has not been fully explored. We developed ECG-based machine learning models to predict risk of mortality among patients presenting to an emergency department or hospital for any reason. Using the 12-lead ECG traces and measurements from 1,605,268 ECGs from 748,773 healthcare episodes of 244,077 patients (2007–2020) in Alberta, Canada, we developed and validated ResNet-based Deep Learning (DL) and gradient boosting-based XGBoost (XGB) models to predict 30-day, 1-year, and 5-year mortality. The models for 30-day, 1-year, and 5-year mortality were trained on 146,173, 141,072, and 111,020 patients and evaluated on 97,144, 89,379, and 55,650 patients, respectively. In the evaluation cohort, 7.6%, 17.3%, and 32.9% patients died by 30-days, 1-year, and 5-years, respectively. ResNet models based on ECG traces alone had good-to-excellent performance with area under receiver operating characteristic curve (AUROC) of 0.843 (95% CI: 0.838–0.848), 0.812 (0.808–0.816), and 0.798 (0.792–0.803) for 30-day, 1-year and 5-year prediction, respectively; and were superior to XGB models based on ECG measurements with AUROC of 0.782 (0.776–0.789), 0.784 (0.780–0.788), and 0.746 (0.740–0.751). This study demonstrates the validity of ECG-based DL mortality prediction models at the population-level that can be leveraged for prognostication at point of care.

Abstract Background: Hospital readmissions are one of the costliest challenges facing healthcare systems, but conventional models fail to predict readmissions well. Many existing models use exclusively manually-engineered features, which are labor intensive and dataset-specific. Our objective was to develop and evaluate models to predict hospital readmissions using derived features that are automatically generated from longitudinal data using machine learning techniques. Methods: We studied patients discharged from acute care facilities in 2015 and 2016 in Alberta, Canada, excluding those who were hospitalized to give birth or for a psychiatric condition. We used population-level linked administrative hospital data from 2011 to 2017 to train prediction models using both manually derived features and features generated automatically from observational data. The target value of interest was 30-day all-cause hospital readmissions, with the success of prediction measured using the area under the curve (AUC) statistic. Results: Data from 428,669 patients (62% female, 38% male, 27% 65 years or older) were used for training and evaluating models: 24,974 (5.83%) were readmitted within 30 days of discharge for any reason. Patients were more likely to be readmitted if they utilized hospital care more, had more physician office visits, had more prescriptions, had a chronic condition, or were 65 years old or older. The LACE readmission prediction model had an AUC of 0.66 ± 0.0064 while the machine learning model’s test set AUC was 0.83 ± 0.0045, based on learning a gradient boosting machine on a combination of machine-learned and manually-derived features. Conclusion: Applying a machine learning model to the computer-generated and manual features improved prediction accuracy over the LACE model and a model that used only manually-derived features. Our model can be used to identify high-risk patients, for whom targeted interventions may potentially prevent readmissions.

Abstract We aimed to determine whether early public health interventions in 2020 mitigated the association of sociodemographic and clinical risk factors with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. We conducted a population-based cohort study of all adults in Ontario, Canada who underwent testing for SARS-CoV-2 through December 31, 2020. The outcome was laboratory-confirmed SARS-CoV-2 infection, determined by reverse transcription polymerase chain reaction testing. Adjusted odds ratios (ORs) were determined for sociodemographic and clinical risk factors before and after the first-wave peak of the pandemic to assess for changes in effect sizes. Among 3,167,753 community-dwelling individuals, 142,814 (4.5%) tested positive. The association between age and SARS-CoV-2 infection risk varied over time (P-interaction

Abstract Aims Heart failure with mildly reduced ejection fraction (HFmrEF) is associated with a favourable prognosis compared with heart failure (HF) with reduced ejection fraction (EF). We assessed whether left ventricular ejection fraction (LVEF) trajectory can be used to identify groups of patients with HFmrEF who have different clinical outcomes in a large retrospective study of patients with serial imaging. Methods and results Patients with HF and ≥2 echocardiograms performed ≥6 months apart were included if the LVEF measured 40–49% on the second study. Patients were classified as HFmrEF‐Increasing if LVEF had increased ≥10% (n = 450), HFmrEF‐Decreasing if LVEF had decreased ≥10% (n = 512), or HFmrEF‐Stable if they did not meet other criteria (n = 389). The primary outcome was all‐cause mortality or cardiovascular hospitalization after the second echocardiogram. Associations with time to first event were assessed with multivariable Cox analyses adjusted for age, co‐morbidities, and medications. In total, 1351 patients with HFmrEF (median age 74, 64.2% male) were included with 28.8% exhibiting stable LVEF. During median follow‐up of 15.3 months, the composite outcome occurred in 811 patients. During follow‐up, patients with HFmrEF‐Increasing were less likely to experience the primary outcome [adjusted hazard ratio (HR) 0.72, 95% confidence interval (CI) 0.60–0.88, P

Background: Data are limited data on the prevalence of cardiovascular disease (CVD) and multimorbidity in contemporary cohorts of high-cost users (HCUs) in Canada.We examined the following: (i) the prevalence of CVD, with a comparison of total healthcare costs among HCUs with vs without CVD; (ii) the contribution of other comorbidities to costs among HCUs with CVD; and (iii) the trajectory of healthcare costs in the years before and after becoming an HCU. Methods: The study included adult Alberta patients in the Canadian Institutes of Health Research/Canadian Institute for Health Information Dynamic Cohort of Complex, High System Users from 2011-2012 through 2014-2015. We examined total healthcare costs, including hospital, ambulatory care, physician services, and drugs. Results: Among 88,536 HCUs, 23.4% had no CVD, 28.9% were hospitalized with a primary diagnosis of CVD, and 47.7% were hospitalized with a secondary diagnosis of CVD. Total healthcare costs were $2.0 billion (20.4% non-hospital costs), $2.8 billion (24.1% non-hospital costs), and $4.9 billion (19.8% non-hospital costs), respectively, in the 3 groups. Many HCUs with CVD were frail (74.2%) and many had diabetes (33.8%) or chronic obstructive pulmonary disease (27.9%), which contributed to higher costs and mortality. Healthcare expenditures in HCUs with CVD were several times higher than per capita health expenditures in the years prior to, and following, their inclusion in the dynamic HCU cohort. Conclusions: CVD is very common in HCUs of healthcare. HCUs with CVD have high rates of frailty and multimorbidity. Further research is needed to identify and intervene earlier, in order to flatten the cost curve in these complex patients. Résumé: Introduction: Les données sur la prévalence des maladies cardiovasculaires (MCV) et de la multimorbidité au sein des cohortes contemporaines de grands utilisateurs (GU) du Canada sont limitées. Nous avons examiné ce qui suit : (i) la prévalence des MCV en comparant les coûts totaux des soins de santé entre les GU atteints de MCV et les GU non atteints de MCV; (ii) la contribution des autres comorbidités aux coûts liés aux GU atteints de MCV; (iii) la trajectoire des coûts des soins de santé dans les années avant et après avoir été considérés comme un GU. Méthodes: L’étude portait sur des patients adultes de l’Alberta de la Cohorte dynamique de grands utilisateurs du système de santé aux besoins complexes de 2011-2012 à 2014-2015 des Instituts de recherche en santé du Canada et de l’Institut canadien d’information sur la santé. Nous avons examiné les coûts totaux des soins de santé, notamment les coûts hospitaliers, les coûts des soins ambulatoires, des services médicaux et des médicaments. Résultats: Parmi les 88 536 GU, 23,4 % n’avaient pas de MCV, 28,9 % étaient hospitalisés et avaient un diagnostic principal de MCV, et 47,7 % étaient hospitalisés et avaient un diagnostic secondaire de MCV. Les coûts totaux des soins de santé des 3 groupes étaient respectivement de 2,0 G$ (20,4 % de coûts non hospitaliers), 2,8 G$ (24,1 % de coûts non hospitaliers) et 4,9 G$ (19,8 % de coûts non hospitaliers). Plusieurs GU atteints de MCV étaient fragiles (74,2 %) et beaucoup avaient le diabète (33,8 %) ou une maladie pulmonaire obstructive chronique (27,9 %), qui contribuaient à des coûts et à une mortalité plus élevés. Les dépenses de santé par personne liées aux GU atteints de MCV étaient beaucoup plus élevées que les dépenses de santé par personne dans les années qui précédaient ou suivaient leur inclusion dans la cohorte dynamique de GU atteints de MCV. Conclusions: Les GU de soins de santé sont très fréquemment atteints de MCV. Les GU atteints de MCV présentent des taux de fragilité et de multimorbidité élevés. D’autres recherches sont nécessaires pour cerner et intervenir plus tôt afin d’aplatir la courbe des coûts chez ces patients aux besoins complexes.

Background Because the impact of changes in how outpatient care was delivered during the COVID‐19 pandemic is uncertain, we designed this study to examine the frequency and type of outpatient visits between March 1, 2019 to February 29, 2020 (prepandemic) and from March 1, 2020 to February 28, 2021 (pandemic) and specifically compared outcomes after virtual versus in‐person outpatient visits during the pandemic. Methods and Results Population‐based retrospective cohort study of all 3.8 million adults in Alberta, Canada. We examined all physician visits and 30‐ and 90‐day outcomes, with a focus on those adults with the cardiovascular ambulatory‐care sensitive conditions heart failure, hypertension, and diabetes. Our primary outcome was emergency department visit or hospitalization, evaluated using survival analysis accounting for competing risk of death. Although in‐person outpatient visits decreased by 38.9% in the year after March 1, 2020 (10 142 184 versus 16 592 599 in the prior year), the introduction of virtual visits (7 152 147; 41.4% of total) meant that total outpatient visits increased by 4.1% in the first year of the pandemic for Albertan adults. Outpatient visit frequency (albeit 41.4% virtual, 58.6% in‐person) and prescribing patterns were stable in the first year after pandemic onset for patients with the cardiovascular ambulatory‐care sensitive conditions we examined, but laboratory test frequency declined by 20% (serum creatinine) to 47% (glycosylated hemoglobin). In the first year of the pandemic, virtual outpatient visits were associated with fewer subsequent emergency department visits or hospitalizations (compared with in‐person visits) for patients with heart failure (adjusted hazard ratio [aHR], 0.90 [95% CI, 0.85–0.96] at 30 days and 0.96 [95% CI, 0.92–1.00] at 90 days), hypertension (aHR, 0.88 [95% CI, 0.85–0.91] and 0.93 [95% CI, 0.91–0.95] at 30 and 90 days), or diabetes (aHR, 0.90 [95% CI, 0.87–0.93] and 0.93 [95% CI, 0.91–0.95] at 30 and 90 days). Conclusions The adoption and rapid uptake of virtual outpatient care during the COVID‐19 pandemic did not negatively impact frequency of follow‐up, prescribing, or short‐term outcomes, and could have potentially positively impacted some of these for adults with heart failure, diabetes, or hypertension in a setting where there was an active reimbursement policy for virtual visits. Given declines in laboratory monitoring and screening activities, further research is needed to evaluate whether long‐term outcomes will differ.

Abstract Aims This study aims to assess long‐term changes in left ventricular ejection fraction (LVEF) together with echocardiographic markers of cardiac remodelling and their association with prognosis and patient‐reported quality of life (QoL). Methods and results We conducted a retrospective analysis of serial echocardiograms performed between January 2009 and December 2019 in 1089 patients (median age 63 years, 71.0% men) enrolled in the Mazankowski Heart Function Clinic Registry who had at least two echocardiograms separated by ≥12 months. We classified the patients into four subgroups by their baseline and LVEF trajectories: persistent heart failure with reduced ejection fraction (persistent HFrEF, n = 364), recovered ejection fraction (HFrecEF, n = 325), transient recovery in ejection fraction (HFtrecEF, n = 117), and preserved ejection fraction (HFpEF, n = 283); 4490 echocardiograms were included in the present analysis, with 4.1 ± 1.8 echocardiograms available per patient during follow‐up. Reductions in echocardiographic markers of cardiac remodelling, including LVIDd [adjusted odds ratio (aOR): 2.22, 95% confidence interval (CI) 1.75–2.86], LVIDs (aOR: 2.44, 95% CI 2.00–2.94), left ventricular mass index (aOR: 1.15, 95% CI 1.09–1.22), E/e′ ratio (aOR: 1.15, 95% CI 1.02–1.30), left atrial volume index (aOR: 1.10, 95% CI 1.03–1.16), along with an increase in the maximum recommended daily dose of renin‐angiotensin system inhibitors (aOR: 1.04, 95% CI 1.01–1.07) and mineralocorticoid‐receptor antagonists (aOR: 1.06, 95% CI 1.01–1.11) at 2 years, strongly predicted the HFrecEF classification, which was further sustained at 5 years of follow‐up. However, changes in these parameters were mostly absent in patients experiencing only a transient recovery in LVEF (HFtrecEF), closely resembling patients with persistent HFrEF. In the multivariable analysis, HFrecEF patients had lower risk of all‐cause mortality alone [adjusted hazard ratio (aHR): 0.46, 95% CI 0.23–0.93], and composite all‐cause (aHR: 0.59, 95% CI 0.49–0.73), cardiovascular (aHR: 0.47, 95% CI 0.36–0.61), and heart failure (aHR: 0.50, 95% CI 0.35–0.70) related hospitalizations with mortality than patients with persistent HFrEF. QoL assessed through the shortened Kansas City Cardiomyopathy Questionnaire‐12 at the end of follow‐up was greater in patients with HFrecEF by 5.2, 12.4, and 9.4 points than persistent HFrEF, HFtrecEF, and HFpEF, respectively. Conclusions Patients with HFrecEF experienced progressive normalization in echocardiographic markers of cardiac remodelling characterized by reductions in left ventricular dimensions and mass in tandem with reductions in left atrial volume and E/e′ ratio, which is associated with better prognosis and QoL.

Background Oral anticoagulation (OAC) therapy prevents morbidity and mortality in nonvalvular atrial fibrillation; whether location of diagnosis influences OAC uptake or adherence is unknown. Methods and Results Retrospective cohort study (2008–2019), identifying adults with incident nonvalvular atrial fibrillation across health care settings (emergency department, hospital, outpatient) at high risk of stroke. OAC uptake and adherence via proportion of days covered for direct OACs and time in therapeutic range for warfarin were measured. Proportion of days covered was categorized as low (0–39%), intermediate (40–79%), and high (80–100%). Warfarin control was defined as time in therapeutic range ≥65%. All‐cause mortality was examined at a 3‐year landmark. Among 75 389 patients with nonvalvular atrial fibrillation (47.0% women, mean 77.4 years), 19.7% were diagnosed in the emergency department, 59.1% in the hospital, and 21.2% in the outpatient setting. Ninety‐day OAC uptake was 51.6% in the emergency department, 50.9% in the hospital, and 67.9% in the outpatient setting (P

Background: Atrial fibrillation and flutter (AFF) are common arrhythmias diagnosed in the emergency department (ED), and prompt follow-up with specialists may yield better outcomes. This study examines time to first specialist outpatient visit following ED discharge for AFF. Methods: Alberta residents aged ≥ 35 years with ED presentations for AFF ending in discharge during 2017-2018 were extracted and linked with hospitalizations and physician claims. A spatial scan and multinomial logistic regression were performed. Regression model predictors included demographics, prior diagnoses, and prior health service use. Results: ED presentations for 4387 patients (54% male; mean age 68 years) were analyzed. Two geographic areas were identified as clusters that had longer times than would be expected by chance: a north cluster of northern areas with an estimated median time of 98 days (95% confidence interval [CI] 82,139), and an east cluster of eastern areas with a median of 57 days (95% CI 47, 68). Patients in the north cluster were more likely to be younger (adjusted odds ratio [aOR] = 0.76 per 5 years, 95% CI 0.62, 0.93) and have prior histories of AFF (aOR = 1.45, 95% CI 1.11, 1.90), congestive heart failure (aOR=1.51, 95% CI 1.15, 1.98), chronic obstructive pulmonary disease (aOR = 2.03, 95% CI 1.55, 2.65), and diabetes (aOR = 1.30, 95% CI 1.00, 1.67). They were less likely to have prior general practitioner outpatient visits (aOR = 0.65 per 5 visits, 95% CI 0.53, 0.81) and specialist outpatient visits (aOR = 0.39, 95% CI 0.30, 0.50) than other patients. Conclusions: Despite being at higher risk, patients in northern areas took longer to see a specialist after an ED presentation for AFF than those from other regions. Innovative strategies for promoting specialist follow-up should be explored. Résumé: Contexte: La fibrillation auriculaire (FA) et le flutter auriculaire sont des troubles arythmiques couramment diagnostiqués dans les services des urgences, et un suivi rapide par un spécialiste pourrait améliorer les résultats pour le patient. La présente étude examine l’intervalle entre le moment où le patient atteint de FA ou de flutter auriculaire obtient son congé du service des urgences et celui de la première consultation externe auprès d’un spécialiste. Méthodologie: Les dossiers des patients albertains âgés de 35 ans ou plus qui se sont présentés au service des urgences pour cause de FA ou de flutter auriculaire et qui ont reçu leur congé en 2017-2018 ont été extraits et couplés aux données de facturation des hôpitaux et des médecins. Une analyse spatiale et une régression logistique multinomiale ont ensuite été réalisées. Le modèle de régression comprenait les facteurs de prédiction suivants : caractéristiques démographiques des patients, diagnostics antérieurs et utilisation antérieure des services de santé. Résultats: Les dossiers de consultation au service des urgences de 4 387 patients (proportion d’hommes : 54 %; âge moyen : 68 ans) ont été analysés. Deux zones géographiques sont ressorties comme des grappes où l’intervalle était plus long que celui auquel on pourrait s’attendre du fait du hasard, soit une grappe de régions du nord où le temps d’attente médian a été estimé à 98 jours (intervalle de confiance [IC] à 95 % : de 82 à 139), et une grappe de régions de l’est où le temps d’attente médian a été estimé à 57 jours (IC à 95 % : de 47 à 68). Les patients de la grappe nord étaient plus susceptibles d’être plus jeunes (risque relatif approché corrigé [RRAc] = 0,76 par tranche de 5 ans; IC à 95 % : de 0,62 à 0,93) et d’avoir des antécédents de FA ou de flutter auriculaire (RRAc = 1,45; IC à 95 % : de 1,11 à 1,90), d’insuffisance cardiaque congestive (RRAc = 1,51; IC à 95 % : de 1,15 à 1,98), de maladie pulmonaire obstructive chronique (RRAc = 2,03; IC à 95 % : de 1,55 à 2,65) et de diabète (RRAc = 1,30; IC à 95 % : de 1,00 à 1,67). Ils étaient moins susceptibles que les autres patients d’avoir déjà consulté en externe un omnipraticien (RRAc = 0,65 par tranche de 5 visites; IC à 95 % : de 0,53 à 0,81) et un spécialiste (RRAc = 0,39; IC à 95 % : de 0,30 à 0,50). Conclusions: Malgré le fait qu’ils soient exposés à un risque plus élevé, les patients de la grappe nord ont attendu plus longtemps que ceux des autres régions avant de consulter un spécialiste après s’être présentés au service des urgences pour cause de FA ou de flutter auriculaire. Il conviendrait d’explorer de nouvelles stratégies pour promouvoir l’importance du suivi par un spécialiste.

Summary: Background: SARS-Cov-2 infection rates are high among residents of long-term care (LTC) homes. We used machine learning to identify resident and community characteristics predictive of SARS-Cov-2 infection. Methods: We linked 26 population-based health and administrative databases to identify the population of all LTC residents tested for SARS-Cov-2 infection in Ontario, Canada. Using ensemble-based algorithms, we examined 484 factors, including individual-level demographics, healthcare use, comorbidities, functional status, and laboratory results; and community-level characteristics to identify factors predictive of infection. Analyses were performed separately for January to April (early wave 1) and May to August (late wave 1). Findings: Among 80,784 LTC residents, 64,757 (80.2%) were tested for SARS-Cov-2 (median age 86 (78–91) years, 30.6% male), of whom 10.2% of 33,519 and 5.2% of 31,238 tested positive in early and late wave 1, respectively. In the late phase (when restriction of visitors, closure of communal spaces, and universal masking in LTC were routine), regional-level characteristics comprised 33 of the top 50 factors associated with testing positive, while laboratory values and comorbidities were also predictive. The c-index of the final model was 0.934, and sensitivity was 0.887. In the highest versus lowest risk quartiles, the odds ratio for infection was 114.3 (95% CI 38.6–557.3). LTC-related geographic variations existed in the distribution of observed infection rates and the proportion of residents at highest risk. Interpretation: Machine learning informed evaluation of predicted and observed risks of SARS-CoV-2 infection at the resident and LTC levels, and may inform initiatives to improve care quality in this setting. Funding: Funded by a Canadian Institutes of Health Research, COVID-19 Rapid Research Funding Opportunity grant (# VR4 172736) and a Peter Munk Cardiac Centre Innovation Grant. Dr. D. Lee is the Ted Rogers Chair in Heart Function Outcomes, University Health Network, University of Toronto. Dr. Austin is supported by a Mid-Career investigator award from the Heart and Stroke Foundation. Dr. McAlister is supported by an Alberta Health Services Chair in Cardiovascular Outcomes Research. Dr. Kaul is the CIHR Sex and Gender Science Chair and the Heart & Stroke Chair in Cardiovascular Research. Dr. Rochon holds the RTO/ERO Chair in Geriatric Medicine from the University of Toronto. Dr. B. Wang holds a CIFAR AI chair at the Vector Institute.

Background Small observational studies have suggested that statin users have a lower risk of dying with COVID‐19. We tested this hypothesis in a large, population‐based cohort of adults in 2 of Canada’s most populous provinces: Ontario and Alberta. Methods and Results We examined reverse transcriptase–polymerase chain reaction swab positivity rates for SARS‐CoV‐2 in adults using statins compared with nonusers. In patients with SARS‐CoV‐2 infection, we compared 30‐day risk of all‐cause emergency department visit, hospitalization, intensive care unit admission, or death in statin users versus nonusers, adjusting for baseline differences in demographics, clinical comorbidities, and prior health care use, as well as propensity for statin use. Between January and June 2020, 2.4% of 226 142 tested individuals aged 18 to 65 years, 2.7% of 88 387 people aged 66 to 75 years, and 4.1% of 154 950 people older than 75 years had a positive reverse transcriptase–polymerase chain reaction swab for SARS‐CoV‐2. Compared with 353 878 nonusers, the 115 871 statin users were more likely to test positive for SARS‐CoV‐2 (3.6% versus 2.8%, P

Abstract Aims Most patients with acute heart failure (AHF) are treated with supplemental oxygen during hospitalization. In this study, we investigated the effect of oxygen titrated to high vs. low pulse oximetry targets in patients hospitalized with AHF. Methods and results In a pilot, open‐label randomized controlled trial (RCT), 50 patients who were admitted with AHF were randomized to either high (≥96%) or low (90–92%) SpO2 targets. Oxygen was manually titrated to the assigned target ranges for 72 h. The primary endpoint was the change in N‐terminal pro‐brain‐type natriuretic peptide (NT‐proBNP) from randomization to 72 h, and secondary endpoints included patient‐reported dyspnoea by visual analogue scale (VAS), patient global assessment (PGA), peak expiratory flow (PEF) within 72 h, and clinical outcomes up to 30 days following hospital discharge. The median age was 73.5 years, and 42% were women. The change in NT‐proBNP was −6963 (−13 345, −1253) pg/mL in the high SpO2 group and −2093 (−5692, −353) pg/mL in the low SpO2 group (P = 0.46), and the 72 h to baseline NT‐proBNP ratio was similar between groups (0.7 vs. 0.6, P = 0.51). There were no differences between arms in change in dyspnoea VAS (P = 0.86), PGA (P = 0.91), PEF (P = 0.52), in‐hospital mortality (4.0% vs. 8.0%, P = 0.50), or 30 day heart failure readmission rates (20.8% vs. 8.7%, P = 0.22). Conclusions In this study, no differences were observed in the primary or secondary outcomes for patients randomized to high vs. low SpO2 targets. Further RCTs with larger sample sizes are warranted to determine the efficacy and safety of oxygen therapy in patients with AHF.

Background: Patients undergoing percutaneous coronary intervention (PCI) are increasingly older and have a higher comorbidity burden. This study evaluated trends in 30-day, 1-year, and 2-year total and cause-specific mortality using a large, contemporary cohort of patients who underwent PCI in Alberta, Canada. Methods: We used the Alberta Provincial Project for Outcome Assessment in Coronary Heart Disease (APPROACH) registry to identify patients aged ≥ 20 years who underwent PCI between 2005 and 2013. All patients were followed until death or being censored by August 2016. Cause of death was from the Vital Statistics database and classified as cardiac or noncardiac. Multivariable logistic regression was used to calculate predicted mortality at 30 days, 1 year, and 2 years post-PCI. Results: Of the 35,602 patients who underwent PCI, 5284 (14.8%) had died. Mean (standard deviation) follow-up was 74.9 (35.1) months. Over the study period, patients were older and more likely to undergo PCI for an acute coronary syndrome indication. Thirty-day (2005: 1.3%; 2013: 3.2%; P < 0.001), 1-year (2005: 2.7%; 2013: 5.7%; P < 0.001), and 2-year (2005: 4.5%; 2013: 7.5%; P < 0.001) predicted mortality after PCI increased over the study period. Cardiac cause of death dominated in the short-term, but the proportion of noncardiac deaths increased as time from PCI to death increased (30 days = 11.5%, 1 year = 31.5%, 2 years = 39.6%; P < 0.001). Conclusions: In this population-based study, we found all-cause mortality at 30 days, 1 year, and 2 years after PCI increased over time. Cardiac causes of death dominate in the short-term after PCI; however, noncardiac cause becomes a major driver of mortality in the long-term. Résumé: Contexte: Les patients devant subir une intervention coronarienne percutanée (ICP) sont de plus en plus âgés et subissent un fardeau accru de comorbidités. La présente étude a évalué les tendances de la mortalité totale et due à une cause particulière à 30 jours, 1 an et 2 ans, au sein d’une vaste cohorte contemporaine de patients ayant subi une ICP en Alberta, au Canada. Méthodologie: Nous avons utilisé le registre APPROACH (Alberta Provincial Project for Outcome Assessment in Coronary Heart Disease) pour recenser les patients âgés de 20 ans ou plus qui avaient subi une ICP entre 2005 et 2013. Tous les patients ont fait l’objet d’un suivi jusqu’au décès ou à la censure des données en août 2016. La cause du décès était issue de la Base de données sur l’état civil et classée comme étant d’origine cardiaque ou non cardiaque. On a eu recours à un modèle de régression logistique multivarié pour calculer la mortalité prédite 30 jours, 1 an et 2 ans après l’ICP. Résultats: Sur les 35 602 patients ayant subi une ICP, 5 284 (14,8 %) étaient décédés. La durée moyenne de suivi (écart type) était de 74,9 (35,1) mois. Au cours de la période de l’étude, les patients étaient plus âgés et plus susceptibles de subir une ICP pour une indication de syndrome coronarien aigu. On observe une augmentation de la mortalité prédite après l’ICP au cours de la période de l’étude selon les taux suivants : à trente jours (2005 : 1,3 %; 2013 : 3,2 %; p < 0,001), à 1 an (2005 : 2,7 %; 2013 : 5,7 %; p < 0,001) et à 2 ans (2005 : 4,5 %; 2013 : 7,5 %; p < 0,001). Les causes cardiaques de décès dominaient à court terme, mais la proportion de décès d’origine non cardiaque augmentait avec le temps au fur et à mesure de l’allongement de l’intervalle entre la date de l’ICP et le décès (30 jours = 11,5 %, 1 an = 31,5 %, 2 ans = 39,6 %; p < 0,001). Conclusions: Dans cette étude de population, nous avons trouvé que la mortalité toutes causes confondues à 30 jours, 1 an et 2 ans après une ICP augmente au fil du temps. Les causes cardiaques de décès dominent peu de temps après l’ICP, tandis que les causes non cardiaques jouent un rôle déterminant dans la mortalité à long terme.

Background: Emergency Heart Failure Mortality Risk Grade (EHMRG) assesses the risk of death within 7 days of emergency department (ED) presentation for patients with acute heart failure (AHF). We aimed to externally validate and refine the EHMRG model in patients who presented to the ED with AHF. Methods: We performed a cohort study using administrative data for all ambulance-transported patients from Alberta (2012-2016) presenting to the ED with a primary diagnosis of AHF. Results: Among 6708 patients with AHF, the 7-day mortality was 0.0%, 0.8%, 1.6%, 4.0%, 4.2%, and 12.0% across EHMRG risk categories (1-4, 5A and 5B). The EHMRG score had a c-index of 0.73 (95% confidence interval [CI], 0.71-0.76) for 7-day mortality and 0.71 (95% CI, 0.70-0.73) for 30-day mortality, but lower c-statistics for other outcomes (0.61-0.67). The inclusion of natriuretic peptides to the EHMRG model improved prediction (Net Reclassification Improvement, 0.268; 95% CI, 0.173-0.363; P < 0.01) for 7-day mortality, as did the addition of the Canadian Triage and Acuity Scale (Net Reclassification Improvement, 0.111; 95% CI, 0.005-0.218; P = 0.04). Conclusion: The EHMRG model exhibited moderate discriminative ability in a large population-based cohort of patients with AHF in the ED. Revision of the EHMRG score through factor inclusion and exclusion could improve the model’s performance. Résumé: Introduction: Le Emergency Heart Failure Mortality Risk Grade (EHMRG) permet d’évaluer le risque de mortalité dans les 7 jours qui suivent la consultation au service des urgences (SU) des patients souffrant d’insuffisance cardiaque aiguë (ICA). Nous avions pour objectif de valider et d’affiner à l’externe le modèle EHMRG chez les patients qui se présentaient au SU en raison d’une ICA. Méthodes: Nous avons réalisé une étude de cohorte à partir des données administratives de tous les patients de l’Alberta transportés par ambulance (2012-2016) au SU en raison d’un diagnostic principal d’ICA. Résultats: Parmi les 6708 patients souffrant d’ICA, la mortalité dans les 7 jours était de 0,0 %, de 0,8 %, de 1,6 %, de 4,0 %, de 4,2 % et de 12,0 % dans toutes les catégories de risque EHMRG (1-4, 5A et 5B). Le score EHMRG montrait un C-index (indice de concordance) de 0,73 (intervalle de confiance [IC] à 95 %, 0,71-0,76) pour la mortalité dans les 7 jours et de 0,71 (IC à 95 %, 0,70-0,73) pour la mortalité dans les 30 jours, mais des statistiques C plus faibles pour les autres résultats cliniques (0,61-0,67). L’inclusion de peptides natriurétiques au modèle EHMRG améliorait la prédiction (indice Net Reclassification Improvement [NRI], 0,268; IC à 95 %, 0,173-0,363; P < 0,01) pour la mortalité dans les 7 jours, de la même manière que l’ajout de l’Échelle canadienne de triage et de gravité (indice NRI, 0,111; IC à 95 %, 0,005-0,218; P = 0,04). Conclusion: Le modèle EHMRG a montré une capacité discriminative modérée dans une vaste cohorte en population générale auprès de patients souffrant d’ICA au SU. La révision du score EHMRG par l’inclusion et l’exclusion de facteurs pourraient permettre d’améliorer la performance du modèle.

Abstract Background As more patients at lower cardiovascular (CV) risk are treated with statins, the balance between cardiovascular benefits and the risk of adverse events becomes increasingly important. Methods We did a population-based cohort study (May 1, 2002 to March 30, 2013) using province-wide laboratory and administrative data in Alberta. We studied new statin users aged 66 years of age and older who were not receiving renal replacement therapy at baseline. We assessed statin use at 30-day intervals to allow time-varying assessment of statin exposure in Cox proportional hazards models that examined the relation between statin use and hospitalization with acute kidney injury (AKI). Results Of the 128,140 new statin users, 47 and 46% were prescribed high- and medium-intensity regimens at the index date. During median follow-up of 4.6 years (interquartile range 2.2, 7.4), 9118 individuals were hospitalized for AKI. Compared to non-use, the use of high- and medium-intensity statin regimens was associated with significant increases in the adjusted risks of hospitalization with AKI: hazard ratios 1.16 [95% confidence interval (CI) 1.10, 1.23] and 1.07 (95% CI 1.01, 1.13), respectively. Risks of AKI were higher among women than men, and among users of angiotensin converting enzyme inhibitors/angiotensin receptor blockers than non-users, and among diuretic users (p for interaction 0.002, 0.01, and 0.04 respectively). Conclusions We found a graded, independent association between the intensity of statin use and the risk of hospitalization with AKI, although the absolute magnitude of the excess risk was small.

Background Improving health‐related quality of life is an important goal in the management of patients with heart failure (HF). Defining health‐related quality of life changes over time in patients with HF with preserved (HFpEF) or reduced ejection fraction and showing their association with other important clinical events could support the use of health‐related quality of life as a measure of quantifying HF care. Methods and Results In the Alberta HEART (Heart Failure Aetiology and Analysis Team) cohort (n=621), patients were categorized into 4 subgroups: healthy controls (n=98), at risk (n=163), HFpEF (n=191), and HF with reduced ejection fraction (n=169). The change of the Kansas City Cardiomyopathy Questionnaire (KCCQ), EuroQOL 5 dimensions, and Functional Assessment of Cancer Therapy—Anemia over 12 months, and its association with a composite of death or rehospitalization within 3 years were assessed. At baseline, the KCCQ overall summary score was 73 (interquartile range, 53–86) in HFpEF and 78 (interquartile range, 56–90) in HF with reduced ejection fraction (P=0.22). Overall, 30.5% of patients with HF experienced ≥5‐point improvements and 32.4% had ≥5‐point worsening in KCCQ overall summary score at 12 months, which did not differ between HFpEF and HF with reduced ejection fraction (P=0.23). Clinical events were higher in patients with HF who had a decline in KCCQ over 12 months as compared with those with stable KCCQ scores (70.2% versus 52.0%, P=0.012). The results were similar for the Functional Assessment of Cancer Therapy—Anemia and EuroQOL 5 dimensions. Conclusions In patients with HF, the KCCQ quantified clinically meaningful changes over time, which were associated with important clinical outcomes in patients with HFpEF. Given the observed variability and prognostication in different patient trajectories, health‐related quality of life measures could be valuable for quantifying the quality of care in healthcare systems.

Abstract Using primary care electronic medical records (the United Kingdom Health Improvement Network Database 2003‐2015), we examined the control of cardiovascular risk factors in the first year after diagnosis in British adults with diabetes mellitus. Among 292 170 individuals with diabetes receiving frequent outpatient management (median of 16 primary care visits in the prior year), control of cardiovascular risk factors a median of 354 days after diagnosis was suboptimal: 14.7% had HbA1C

Abstract Background To examine the intensity of glycemic and blood pressure control in British adults with diabetes mellitus and whether control levels or treatment deintensification rates differ across health states. Methods Retrospective cohort study using primary care electronic medical records (the United Kingdom Health Improvement Network Database) for adults with diabetes diagnosed at least 6 months before the index HbA1C and systolic blood pressure (SBP) measurements (to give their primary care physicians time to achieve treatment goals). We used prescribing records for 6 months pre/post the index measurements to determine who had therapy subsequently deintensified (based on “glycemic therapy score” and “antihypertensive therapy score” derived from number and dosage of medications). Results Of 292,170 individuals with diabetes, HbA1C 7.5% and 46.6, 51.4, and 48.5% had SBP > 140 mmHg). The proportions of patients with HbA1C or SBP out of recommended treatment ranges changed little 6 months later despite frequent (median 14 per year) primary care visits. Conclusions Glycemic and hypertensive control exhibited statistically significant but small magnitude differences across frailty states. Medication deintensification was uncommon, even in frail patients below SBP and HbA1C targets. SBP levels were more likely to be outside recommended treatment ranges than glycemic levels. Trial registration As this study is a retrospective secondary analysis of electronic medical record data and not a health care intervention trial it was not registered

Abstract Background To assess the use of updated comorbidity information over time on ability to predict mortality among adults with newly diagnosed hypertension. Methods We studied adults 18 years and older with an incident diagnosis of hypertension from Alberta, Canada. We compared the prognostic performance of Cox regression models using Charlson comorbidities as time-invariant covariates at baseline (TIC) versus models including Charlson comorbidities as time-varying covariates (TVC) using Akaike Information Criterion (AIC) for testing goodness of fit. Results The strength of the association between important prognostic clinical variables and mortality varied by modeling technique; for example, myocardial infarction was less strongly associated with mortality in the TIC model (Hazard Ratio 1.07; 95% Confidence Interval (CI): 1.05 to 1.1) than in the TVC model (HR 1.20; 95% CI: 1.18 to 1.22). All TVC models slightly outperformed TIC models, regardless of the method used to adjust for comorbid conditions (individual Charlson Comorbidities, count of comorbidities or indices). The TVC model including all 17 Charlson comorbidities as individual independent variables showed the best fit and performance. Conclusion Accounting for changes in patient comorbidity status over time more accurately captures a patient’s health risk and slightly improves predictive model fit and performance than traditional methods using TIC assessment.

BackgroundAs questions have been raised about the appropriateness of direct oral anticoagulant (DOAC) dosing among outpatients with atrial fibrillation, we examined this issue in patients being managed by primary care providers. Methods and ResultsThis was a retrospective cohort new‐user study using electronic medical records from 744 Canadian primary care clinicians. Potentially inappropriate DOAC prescribing was defined as prescribing lower or higher doses than those recommended by guidelines for patients with nonvalvular atrial fibrillation. Of the 6658 patients with nonvalvular atrial fibrillation who were prescribed a DOAC (mean age: 74.8; 55% male), 626 (9.4%) had a CHADS2 score of 0, and 168 (2.5%) had a CHADS‐VASc score of 0. Of the DOAC prescriptions, 527 (7.7%) were deemed potentially inappropriate: 496 (7.2%) were potentially underdosed, and 31 (0.5%) were prescribed a dose that was higher than recommended. Patients were more likely to be prescribed lower‐than‐recommended doses if they were female (adjusted odds ratio [aOR]: 1.3 [95% confidence interval (CI), 1.0–1.5]), had multiple comorbidities (aOR: 1.4 [95% CI, 1.1–1.8])—particularly heart failure (aOR: 1.6 [95% CI, 1.2–2.0]) or dementia (aOR: 1.4 [95% CI, 1.1–1.8])—or if they were also taking aspirin (aOR: 1.7 [95% CI, 1.3–2.1]) or nonsteroidal anti‐inflammatory drugs (aOR: 1.2 [95% CI, 1.02–1.5]). Potentially inappropriate DOAC dosing was more common in rural practices (aOR: 2.1 [95% CI, 1.7–2.6]) or smaller practices (aOR: 1.9 [95% CI, 1.6–2.4] for practices smaller than median). ConclusionsThe vast majority of DOAC prescriptions in our cohort of primary care–managed patients appeared to be for appropriate doses, particularly since prescribing a reduced dose of DOAC may be appropriate in frail patients or those taking other medications that predispose to bleeding.

Objectives: Concerns have been raised that isolation precautions may have unintended consequences. The relationship between patient isolation and the 30-day risk of readmission or death among patients discharged from a general medicine ward was examined. Methods: A prospective cohort study of adult patients discharged to the community from seven general internal medicine wards in Edmonton, Alberta, Canada, from October 2013 to November 2014, was performed. Patients under contact, respiratory, or droplet precautions were considered isolated. Covariates measured at discharge included the Charlson comorbidity score, LACE index, clinical frailty, depression, anxiety, health-related quality of life, and patient satisfaction. Outcomes were measured at 30 days by telephone follow-up and provincial electronic health record query. Results: Of 495 patients (mean age 62 years, 51% female), 75 (18%) were isolated during their admission. Isolated and non-isolated patients had similar lengths of stay (6.2 vs. 6.2 days), depression, anxiety, health-related quality of life, and satisfaction scores at discharge (all p-values non-significant). At 30 days, 85 (17.2%) patients had been readmitted or had died (20.0% of isolated patients vs. 16.7% of non-isolated patients; adjusted odds ratio 1.11, 95% confidence interval 0.57–2.18). Conclusions: In-hospital isolation does not appear to have an adverse impact on outcomes once patients are discharged from hospital.

BackgroundUp to one fifth of readmissions after a heart failure hospitalization occur at a different hospital. This negatively impacts information continuity, but whether site of readmission impacts subsequent outcomes is unclear. Methods and ResultsRetrospective cohort study of all patients discharged with a primary diagnosis of heart failure in Canada between April 2004 and December 2013. We compared patients readmitted within 30 days to the original hospital versus a different hospital. Of the 217 039 heart failure patients (mean age, 76.8 years, 50.1% male), 39 368 (18.1%) were readmitted within 30 days—32 771 (83.2%) to the original hospital and 6597 (16.8%) to a different hospital (increasing over time from 15.6% in 2004 to 18.5% by 2013; P for trend=0.001). Patients readmitted to different hospitals were younger and were more likely to be male, have a rural residence, a more‐recent discharge year, an index hospitalization at a teaching hospital, and to be brought in by ambulance at the time of the readmission. Readmissions to the original hospital were substantially shorter (mean, 10.4 days [95% CI, 10.3–10.6] versus 11.6 days [95% CI, 11.3–12.0]; adjusted means, 11.0 versus 12.0; P

AimsPatients visiting the emergency department (ED) or hospitalized for heart failure (HF) are at increased risk for subsequent adverse outcomes, however effective risk stratification remains challenging. We utilized a machine-learning (ML)-based approach to identify HF patients at risk of adverse outcomes after an ED visit or hospitalization using a large regional administrative healthcare data system.Methods and resultsPatients visiting the ED or hospitalized with HF between 2002-2016 in Alberta, Canada were included. Outcomes of interest were 30-day and 1-year HF-related ED visits, HF hospital readmission or all-cause mortality. We applied a feature extraction method using deep feature synthesis from multiple sources of health data and compared performance of a gradient boosting algorithm (CatBoost) with logistic regression modelling. The area under receiver operating characteristic curve (AUC-ROC) was used to assess model performance. We included 50,630 patients with 93,552 HF ED visits/hospitalizations. At 30-day follow-up in the holdout validation cohort, the AUC-ROC for the combined endpoint of HF ED visit, HF hospital readmission or death for the Catboost and logistic regression models was 74.16 (73.18-75.11) versus 62.25 (61.25-63.18), respectively. At 1-year follow-up corresponding values were 76.80 (76.1-77.47) versus 69.52 (68.77-70.26), respectively. AUC-ROC values for the endpoint of all-cause death alone at 30-days and 1-year follow-up were 83.21 (81.83-84.41) versus 69.53 (67.98-71.18), and 85.73 (85.14-86.29) versus 69.40 (68.57-70.26), for the CatBoost and logistic regression models, respectively.ConclusionsML-based modelling with deep feature synthesis provided superior risk stratification for HF patients at 30-days and 1-year follow-up after an ED visit or hospitalization using data from a large administrative regional healthcare system.