David Barber

Prevalence and Epidemiology of Diabetes in Canadian Primary Care Practices: A Report from the Canadian Primary Care Sentinel Surveillance Network

OBJECTIVE The Canadian Primary Care Sentinel Surveillance Network (CPCSSN) is a large, validated national primary care Electronic Medical Records (EMR)-based database. Our objective was to describe the epidemiology of diabetes in this Canadian sample. METHODS We analyzed the records of 272 469 patients10 years of age and older, with at least 1 primary care clinical encounter between January 1, 2011, and December 31, 2012. We calculated the age-gender standardized prevalence of diabetes. We compared health care utilization and comorbidities for 7 selected chronic conditions in patients with and without diabetes. We also examined patterns of medication usage. RESULTS The estimated population prevalence of diabetes was 7.6%. Specifically, we studied 25 425 people with diabetes who had at least 1 primary care encounter in 2 years. On average, patients with diabetes had 1.42 times as many practice encounters as patients without diabetes (95% CI 1.42 to 1.43, p<0.0001). Patients with diabetes had 1.29 times as many other comorbid conditions as those without diabetes (95% CI 1.27 to 1.31, p<0.0001). We found that 85.2% of patients taking hypoglycemic medications were taking metformin, and 51.8% were taking 2 or more classes of medications. CONCLUSIONS This study is the first national Canadian report describing the epidemiology of diabetes using primary care EMR-based data. We found significantly higher rates of primary care use, and greater numbers of comorbidities in patients with diabetes. Most patients were on first-line hypoglycemic medications. Data routinely recorded in EMRs can be used for surveillance of chronic diseases such as diabetes in Canada. These results can enable comparisons with other national EMR-based datasets.

The diagnosis of depression and its treatment in Canadian primary care practices: an epidemiological study

BACKGROUND A diagnosis of depression is common in primary care practices, but data are lacking on the prevalence in Canadian practices. We describe the prevalence of the diagnosis among men and women, patient characteristics and drug treatment in patients diagnosed with depression in the primary care setting in Canada. METHODS Using electronic medical record data from the Canadian Primary Care Sentinel Surveillance Network, we examined whether the prevalence of a depression diagnosis varied by patient characteristics, the number of chronic conditions and the presence of the following chronic conditions: hypertension, diabetes, chronic obstructive pulmonary disease, osteoarthritis, dementia, epilepsy and parkinsonism. We used regression models to examine whether patient characteristics and type of comorbidity were associated with a depression diagnosis. RESULTS Of the 304 412 patients who had at least 1 encounter with their primary care provider between Jan. 1, 2011, and Dec. 31, 2012, 14% had a diagnosis of depression. Current or past smokers and women with a high body mass index had higher rates of depression. One in 4 patients with a diagnosis of depression also had another chronic condition; those with depression had 1.5 times more primary care visits. About 85% of patients with depression were prescribed medication, most frequently selective serotonin reuptake inhibitors, followed by atypical antipsychotics. INTERPRETATION Our data provide information on the prevalence of a depression diagnosis in primary care and associations with being female, having a chronic condition, smoking history and obesity in women. Our findings may inform research and assist primary care providers with early detection and interventions in at-risk patient populations.

A Canadian Primary Care Sentinel Surveillance Network Study Evaluating Antidepressant Prescribing in Canada From 2006 to 2012

Objective: To evaluate the prescribing patterns of antidepressants (ADs) by primary care providers to youth, adults, and seniors, from 2006 to 2012, using data from electronic medical records (EMRs). Method: This was a retrospective cross-sectional database study that used primary care data from the Canadian Primary Care Sentinel Surveillance Network (CPCSSN). Data on more than 600 000 Canadian primary care patients were used to determine the prevalence and incidence of AD prescribing to patients 15 years and older who had an encounter in the years of study (from 2006 to 2012). Each study year was evaluated independently. Results: The study population consisted of 86 927 patients in 2006 (mean age 48.1 years [SD 18.7], 38% male) and grew to 273 529 (mean age 49.6 years [SD 19.3], 40% male) in 2012. The prevalence of AD prescribing increased from 9.20% in 2006 to 12.80% in 2012 (P < 0.001). While the incidence rate of AD prescribing dropped from 3.54% in 2006 to 2.72% in 2008 (P < 0.001) the rate started to significantly rise again, reaching an incidence of 3.07% by 2012 (P < 0.001). Conclusions: The prevalence of AD prescribing by primary care providers in Canada continued to rise from 2006 to 2012. Conversely, incidence has remained stable or declined during the 6-year study period. While many complex factors likely contribute to the observed prevalence and incidence rates, our findings suggest that the guidelines indicating the efficacy of long-term AD therapy for patients with highly recurrent or severe depression are being followed.

Increasing Incidence and Prevalence of World Health Organization Groups 1 to 4 Pulmonary Hypertension: A Population-Based Cohort Study in Ontario, Canada

Background: The World Health Organization recognizes 5 groups of pulmonary hypertension (PH), categorized by pathogenesis or comorbidity: 1-pulmonary arterial hypertension 2-left-heart disease, 3-lung disease and hypoxia 4-chronic thromboembolic disease, and 5-miscellaneous. The epidemiology of PH, apart from group 1, is largely unknown. Methods and Results: We describe incidence, prevalence, comorbidities, mortality and prescribing patterns for groups 1 to 4 PH from 1993 to 2012. Case definitions are based on hospitalizations and emergency department visits, using the Institute for Clinical Evaluative Sciences data, which comprises linked databases of universal coverage health service records for Ontario residents. This cohort included 50 529 patients with PH. The annual incidence of adult PH increased from 2003 to 2012 from 24.1 to 28.7 cases/100 000 population and the annual prevalence from 1993 to 2012 from 99.8 to 127.3 cases/100 000 population, respectively. The most common form of adult PH was group 2, alone (34.2%) or combined with group 3 PH (29.3%). A diagnosis of PH increased the 1-year standardized mortality ratio 7.2-fold. Mortality in adults with PH was 13.0%, 36.4%, and 62.4%, at 30 days, 1 year, and 5 years, respectively. Mortality was highest in groups 2 and 3 and lowest in group 1. PH was present in only 3.6% of people with left heart disease, 0.7% with lung disease, and 1.4% with thromboembolic disease, suggesting that PH is a relatively rare complication of these common diseases. Children (age<16 years) accounted for 3.6% of the cohort. In children group 1 PH was most common (65.2%), and 5-year mortality was lower (21.4%) than in adults. Group 1–specific PH therapies were increasingly prescribed over time and paradoxically were often used in patients who seemed to have group 2, PH based on diagnostic codes indicating left heart disease. Conclusions: The incidence and prevalence of adult PH are increasing. Groups 2 and 3 are the most common and lethal forms of PH. This study identifies an emerging epidemic of PH that likely has substantial adverse health and economic implications.

Adult obesity prevalence in primary care users: An exploration using Canadian Primary Care Sentinel Surveillance Network (CPCSSN) data.

ObjectivesThis research examines the feasibility of using electronic medical records within the Canadian Primary Care Sentinel Surveillance Network (CPCSSN) for obesity surveillance in Canada by assessing obesity trends over time and comparing BMI distribution estimates from CPCSSN to those obtained from nationally representative surveys.MethodsData from 2003–2012 on patients 18 years and older (n = 216,075) were extracted from the CPCSSN database. Patient information included demographics (age and sex) and anthropometric measures (height, weight, body mass index (BMI), waist circumference, and waist-to-hip ratio). Standard descriptive statistics were used to characterize the sample, including, as appropriate, means, proportions and medians. The BMI distribution of the CPCSSN population was compared to estimates from the Canadian Community Health Survey (CCHS) and the Canadian Health Measures Survey (CHMS) for the years: 2004, 2007–2009 and 2009–2011.ResultsThe estimated prevalence of obesity increased from 17.9% in 2003 to 30.8% in 2012. Obesity class I, II and III prevalence estimates from CPCSSN in 2009–2011 (18.0%, 95% CI: 17.8–18; 7.4%, 95% CI: 7.3–7.6; 4.2%, 95% CI: 4.1–4.3 respectively) were greater than those from the most recent (2009–2011) cycle of the CHMS (16.2%, 95% CI: 14–18.7; 6.3%, 95% CI: 4.6–8.5; 3.7%, 95% CI: 2.8–4.8 respectively), however these differences were not statistically significant.ConclusionThe data from CPCSSN present a unique opportunity for longitudinal obesity surveillance among primary care users in Canada, and offer prevalence estimates similar to those obtained from nationally representative survey data.RésuméObjectifsNous avons examiné la faisabilité d’utiliser les dossiers médicaux électroniques au sein du Réseau canadien de surveillance sentinelle en soins primaires (RCSSSP) pour la surveillance de l’obésité au Canada en évaluant la progression de l’obésité au fil du temps et en comparant les estimations de répartition de l’IMC du RCSSSP à celles obtenues dans des enquêtes nationales représentatives.MéthodeNous avons extrait de la base de données du RCSSSP les données de 2003–2012 sur les patients de 18 ans et plus (n = 216 075). Les renseignements sur les patients étaient leur profil démographique (âge et sexe) et leurs mesures anthropométriques (taille, poids, indice de masse corporelle [IMC], périmètre ombilical et rapport taille-hanches). Des statistiques descriptives types ont servi à caractériser l’échantillon, notamment, le cas échéant, les moyennes, les proportions et les médianes. La répartition de l’IMC dans la population du RCSSSP a été comparée aux estimations de l’Enquête sur la santé dans les collectivités canadiennes (ESCC) et de l’Enquête canadienne sur les mesures de la santé (ECMS) pour les années 2004, 2007–2009 et 2009–2011.RésultatsLa prévalence estimative de l’obésité est passée de 17,9 % en 2003 à 30,8 % en 2012. Les estimations de la prévalence de l’obésité de classe I, II et III dans la population du RCSSSP en 2009–2011 (18 %, IC de 95 %: 17,8–18; 7,4 %, IC de 95 %: 7,3–7,6; 4,2 %, IC de 95 %: 4,1–4,3, respectivement) étaient supérieures à celles du cycle le plus récent (2009–2011) de l’ECMS (16,2 %, IC de 95 %: 14–18,7; 6,3 %, IC de 95 %: 4,6–8,5; 3,7 %, IC de 95 %: 2,8–4,8, respectivement), mais ces différences n’étaient pas significatives.ConclusionLes données du RCSSSP offrent une occasion unique de faire une surveillance longitudinale de l’obésité chez les utilisateurs de soins primaires au Canada, et elles donnent des estimations de prévalence semblables à celles obtenues par les données d’enquêtes nationales représentatives.

Are We Asking Patients if They Smoke?: Missing Information on Tobacco Use in Canadian Electronic Medical Records.

INTRODUCTION All adolescent and adult patients should be asked if they smoke. Data entered in electronic medical records offer new opportunities to study tobacco-related clinical activities. The purpose of this study is to examine the recording of tobacco use in Canadian electronic medical records. METHODS Data were collected on September 30, 2013, and analyzed in 2014. Data on 249,223 patients that were aged ≥16 years as of September 30, 2013 and had at least one primary care encounter in the previous 2 years were included. The proportion of patients with information on tobacco use entered in a summative health profile was calculated. Associations between data gaps and patient or physician factors were examined. RESULTS Information on tobacco use was available for 64.4% of patients. Physicians using an electronic medical record for ≥4 years were more likely to have data (AOR=4.57, 95% CI=1.84, 7.29, p<0.0001). Patients aged ≥30 years were more likely to have tobacco information present (AOR=2.92, 95% CI=2.82, 3.02, p<0.0001, for patients aged 30-59 compared to those aged <30 years), as were patients with any comorbidities (AOR=1.41, 95% CI=1.36, 1.45, p<0.0001, for patients with one or two comorbidities compared with none) or more visits. CONCLUSIONS A third of Canadians in this sample lacked data on tobacco in their electronic medical record. Younger, healthier people were less likely to have information about their smoking status. Efforts to improve the recording of tobacco-related information in electronic medical records, especially for younger patients, are needed.

Emergency Department Re-Presentation for Atrial Fibrillation and Atrial Flutter.

BACKGROUND Identification and appropriate management of patients with atrial fibrillation (AF) is critical to mitigate the consequences of the disease. We sought to assess the frequency and pattern of the emergency department (ED) use by patients who presented with AF and/or atrial flutter (AFL) in a midsized Canadian hospital. METHODS We conducted a retrospective cohort analysis of patients who presented to the ED with AF and/or AFL during the calendar years 2010-2012. Patients were identified using the MUSE (General Electric Healthcare, Bucks, United Kingdom) electrocardiogram database and matched with the National Ambulatory Care Reporting System and Discharge Abstract Database up to and including December 31, 2013, a follow-up period of 12 months. The number of presentations and time between visits was assessed. Admissions were identified and lengths of stay and reason for admission were recorded. RESULTS We identified 1361 patients who represented a total of 4783 visits to the ED, a mean of 2.8 ± 2.9 visits per patient with 949 (69.7%) who returned for a subsequent ED visit in the subsequent 12 months. Mean time between base and subsequent visits was 136.8 ± 114.2 days. ED visits generated 1462 admissions (63.0% at repeat ED visits); mean length of stay was 9.7 ± 16.0 days. Stroke or transient ischemic attack accounted for 80 return visits and 8 deaths in 77 patients, 74% of whom with subtherapeutic or no anticoagulation medication. CONCLUSIONS Presentation to the ED with AF and/or AFL, either as a primary reason for consultation or as a secondary diagnosis, was associated with a high risk of subsequent re-presentation and hospital admission.

Recent Trends in the Prescribing of ADHD Medications in Canadian Primary Care

Objective: The aim of this study was to describe the prevalence and incidence of ADHD medication prescribing, by age and gender, from 2005 to 2015 in Canadian primary care. Method: A population-based retrospective cohort study was conducted to evaluate the prescribing of ADHD medications between 2005 and 2015 using electronic medical record data. Yearly prevalence and incidence of ADHD medication prescribing were calculated for preschoolers (up to 5 years old), school-aged children (6-17 years old), and adults (18-65 years old) along with a description of the types of ADHD medications prescribed between 2005 and 2015. Results: Between 2005 and 2015, there was a 2.6-fold increase in the prevalence of ADHD medication prescribing to preschoolers, a 2.5-fold increase in school-aged children, and a fourfold increase in adults. There was a corresponding rise in incidence of prescribing although this rise was moderate and estimates were much lower compared with prevalence. The most commonly prescribed medication was Methylphenidate (65.0% of all ADHD medications prescribed). Conclusion: Although the prevalence of ADHD has remained stable over time, this study found an increase in the prescribing of ADHD medications in all age groups between 2005 and 2015. Incidence of new prescriptions was small relative to prevalence, suggesting that longer term treatments are being adopted.

Detecting Low Back Pain from Clinical Narratives Using Machine Learning Approaches

Free-text clinical notes recorded during the patients’ visits in the Electronic Medical Record (EMR) system narrates clinical encounters, often using ‘SOAP’ notes (an acronym for subject, objective, assessment, and plan). The free-text notes represent a wealth of information for discovering insights, particularly in medical conditions such as pain and mental illness, where regular health metrics provide very little knowledge about the patients’ medical situations and reactions to treatments. In this paper, we develop a generic text-mining and decision support framework to diagnose chronic low back pain. The framework utilizes open-source algorithms for anonymization, natural language processing, and machine learning to classify low back pain patterns from unstructured free-text notes in the Electronic Medical Record (EMR) system as noted by the primary care physicians during patients’ visits. The initial results show a high accuracy for the limited thirty-four patient labelled data set that we used in this pilot study. We are currently processing a larger data set to test our approach.

Determining the impact of a new physiotherapist-led primary care model for back pain: protocol for a pilot cluster randomized controlled trial

BackgroundBack pain is a leading contributor to disability, healthcare costs, and lost work. Family physicians are the most common first point of contact in the healthcare system for people with back pain, but physiotherapists (PTs) may be able to support the primary care team through evidence-based primary care. A cluster randomized trial is needed to determine the clinical, health system, and societal impact of a primary care model that integrates physiotherapists at the first visit for people with back pain. Prior to conducting a future fully powered cluster randomized trial, we need to demonstrate feasibility of the methods. Therefore, the purpose of this pilot study will be to:1)Determine feasibility of patient recruitment, assessment procedures, and retention.2)Determine the feasibility of training and implementation of a new PT-led primary care model for low back pain (LBP)3)Explore the perspectives of patients and healthcare providers (HCPs) related to their experiences and attitudes towards the new service delivery model, barriers/facilitators to implementation, perceived satisfaction, perceived value, and impact on clinic processes and patient outcomes.MethodsThis pilot cluster randomized controlled trial will enroll four sites and randomize them to implement a new PT-led primary care model for back pain or a usual physician-led primary care model. All adults booking a primary care visit for back pain will be invited to participate. Feasibility outcomes will include: recruitment and retention rates, completeness of assessment data, PT training participation and confidence after training, and PT treatment fidelity. Secondary outcomes will include the clinical, health system, cost, and process outcomes planned for the future fully powered cluster trial. Results will be analyzed and reported descriptively and qualitatively. To explore perspectives of both HCPs and patients, we will conduct semi-structured qualitative interviews with patients and focus groups with HCPs from participants in the PT-led primary care sites.DiscussionIf this pilot demonstrates feasibility, a fully powered trial will provide evidence that has the potential to transform primary care for back pain. The full trial will inform future service design, whether these models should be more widely implemented, and training agendas.Trial registrationClinicalTrials.gov, NCT03320148. Submitted for registration on 17 September 2017.