Can Patients with Dementia Be Identified in Primary Care Electronic Medical Records Using Natural Language Processing?

Dementia and mild cognitive impairment can be underrecognized in primary care practice and research. Free-text fields in electronic medical records (EMRs) are a rich source of information which might support increased detection and enable a better understanding of populations at risk of dementia. We used natural language processing (NLP) to identify dementia-related features in EMRs and compared the performance of supervised machine learning models to classify patients with dementia. We assembled a cohort of primary care patients aged 66 + years in Ontario, Canada, from EMR notes collected until December 2016: 526 with dementia and 44,148 without dementia. We identified dementia-related features by applying published lists, clinician input, and NLP with word embeddings to free-text progress and consult notes and organized features into thematic groups. Using machine learning models, we compared the performance of features to detect dementia, overall and during time periods relative to dementia case ascertainment in health administrative databases. Over 900 dementia-related features were identified and grouped into eight themes (including symptoms, social, function, cognition). Using notes from all time periods, LASSO had the best performance (F1 score: 77.2%, sensitivity: 71.5%, specificity: 99.8%). Model performance was poor when notes written before case ascertainment were included (F1 score: 14.4%, sensitivity: 8.3%, specificity 99.9%) but improved as later notes were added. While similar models may eventually improve recognition of cognitive issues and dementia in primary care EMRs, our findings suggest that further research is needed to identify which additional EMR components might be useful to promote early detection of dementia. Supplementary Information: The online version contains supplementary material available at 10.1007/s41666-023-00125-6.

Trends in Health Service Use for Canadian Adults With Dementia and Parkinson Disease During the First Wave of the COVID-19 Pandemic.

Importance: Persons with dementia and Parkinson disease (PD) are vulnerable to disruptions in health care and services. Objective: To examine changes in health service use among community-dwelling persons with dementia, persons with PD, and older adults without neurodegenerative disease during the first wave of the COVID-19 pandemic. Design Setting and Participants: Repeated cross-sectional analysis using population-based administrative data among community-dwelling persons with dementia, persons with PD, and adults 65 years and older at the start of each week from March 1 through the week of September 20, 2020 (pandemic period), and March 3 through the week of September 22, 2019 (historical period), in Ontario, Canada. Exposures: COVID-19 pandemic as of March 1, 2020. Main Outcomes and Measures: Main outcomes were weekly rates of emergency department visits, hospitalizations, nursing home admissions, home care, virtual and in-person physician visits, and all-cause mortality. Poisson regression models were used to calculate weekly rate ratios (RRs) with 95% CIs comparing pandemic weeks with historical levels. Results: Among those living in the community as of March 1, 2020, persons with dementia (n = 131 466; mean [SD] age, 80.1 [10.1] years) were older than persons with PD (n = 30 606; 73.7 [10.2] years) and older adults (n = 2 363 742; 74.0 [7.1] years). While all services experienced declines, the largest drops occurred in nursing home admissions (RR for dementia: 0.10; 95% CI, 0.07-0.15; RR for PD: 0.03; 95% CI, 0.00-0.21; RR for older adults: 0.11; 95% CI, 0.06-0.18) and emergency department visits (RR for dementia: 0.45; 95% CI, 0.41-0.48; RR for PD: 0.40; 95% CI, 0.34-0.48; RR for older adults: 0.45; 95% CI, 0.44-0.47). After the first wave, most services returned to historical levels except physician visits, which remained elevated (RR for dementia: 1.07; 95% CI, 1.05-1.09; RR for PD: 1.10, 95% CI, 1.06-1.13) and shifted toward virtual visits. Older adults continued to experience lower hospitalizations. All-cause mortality was elevated across cohorts. Conclusions and Relevance: In this population-based repeated cross-sectional study in Ontario, Canada, those with dementia, those with PD, and older adults sought hospital care far less than usual, were not admitted to nursing homes, and experienced excess mortality during the first wave of the pandemic. Most services returned to historical levels, but virtual physician visits remained a feature of care. While issues of equity and quality of care are still emerging among persons with neurodegenerative diseases, policies to support virtual care are necessary.

Visits to the emergency department by community-dwelling people with dementia during the first 2 waves of the COVID-19 pandemic in Ontario: a repeated cross-sectional analysis.

BACKGROUND: Community-dwelling people with dementia have been affected by COVID-19 pandemic health risks and control measures that resulted in worsened access to health care and service cancellation. One critical access point in health systems is the emergency department. We aimed to determine the change in weekly rates of visits to the emergency department of community-dwelling people with dementia in Ontario during the first 2 waves of the COVID-19 pandemic compared with historical patterns. METHODS: We conducted a population-based repeated cross-sectional study and used health administrative databases to compare rates of visits to the emergency department among community-dwelling people with dementia who were aged 40 years and older in Ontario during the first 2 waves of the COVID-19 pandemic (March 2020-February 2021) with the rates of a historical period (March 2019-February 2020). Weekly rates of visits to the emergency department were evaluated overall, by urgency and by chapter from the International Statistical Classification of Diseases and Related Health Problems, 10th Revision. We used Poisson models to compare pandemic and historical rates at the week of the lowest rate during the pandemic period and the latest week. RESULTS: We observed large immediate declines in rates of visits to the emergency department during the COVID-19 pandemic (rate ratio [RR] 0.50, 95% confidence interval [CI] 0.47-0.53), which remained below historical levels by the end of the second wave (RR 0.88, 95% CI 0.83-0.92). Rates of both nonurgent (RR 0.33, 95% CI 0.28-0.39) and urgent (RR 0.51, 95% CI 0.48-0.55) visits to the emergency department also declined and remained low (RR 0.68, 95% CI 0.59-0.79, RR 0.91, 95% CI 0.86-0.96), respectively. Visits for injuries, and circulatory, respiratory and musculoskeletal diseases declined and remained below historical levels. INTERPRETATION: Prolonged reductions in visits to the emergency department among people with dementia during the first 2 pandemic waves raise concerns about patients who delay seeking acute care services. Understanding the long-term effects of these reductions requires further research.

Impact of the COVID-19 Pandemic on Home Care Services Among Community-Dwelling Adults With Dementia.

OBJECTIVE: To examine how the COVID-19 pandemic impacted use of home care services for individuals with dementia across service types and sociodemographic strata. DESIGN: Population-based time series analysis. SETTING AND PARTICIPANTS: Community-dwelling adults with dementia in Ontario, Canada, from January 2019 to September 2020. METHODS: We used health administrative databases (Ontario Registered Persons Database and Home Care Database) to measure home care services used by participants. Poisson regression models were fit to compare weekly rates of home care services during the pandemic to historical trends with rate ratios (RRs) and 95% confidence intervals (CIs) stratified by service type (nursing, personal care, therapy), sex, rurality, and neighborhood income quintile. RESULTS: During the first wave of the pandemic, personal care fell by 16% compared to historical levels (RR 0.84, 95% CI 0.84, 0.85) and therapies fell by 50% (RR 0.50, 95% CI 0.48, 0.52), whereas nursing did not significantly decline (RR 1.02, 95% CI 1.00, 1.04). All rates had recovered by September 2020, with nursing and therapies higher than historical levels. Changes in services were largely consistent across sociodemographic strata, although the rural population experienced a larger decline in personal care and smaller rebound in nursing. CONCLUSIONS AND IMPLICATIONS: Personal care and therapies for individuals with dementia were interrupted during the early months of the pandemic, whereas nursing was only minimally impacted. Pandemic responses with the potential to disrupt home care for individuals living with dementia must balance the impacts on individuals with dementia, caregivers, and providers.

Validation of canadian health administrative data algorithms for estimating trends in the incidence and prevalence of osteoarthritis.

Objective: To estimate the 1) accuracy of algorithms for identifying osteoarthritis (OA) using health administrative data; and 2) population-level OA prevalence and incidence over time in Ontario, Canada. Method: We performed a retrospective chart abstraction study to identify OA patients in a random sample of 7500 primary care patients from electronic medical records. The validation sample was linked with several administrative data sources. Accuracy of administrative data algorithms for identifying OA was tested against two reference standard definitions by estimating the sensitivity, specificity and predictive values. The validated algorithms were then applied to the Ontario population to estimate and compare population-level prevalence and incidence from 2000 to 2017. Results: OA prevalence within the validation sample ranged from 10% to 23% across the two reference standards. Algorithms varied in accuracy depending on the reference standard, with the sensitivity highest (77%) for patients with OA documented in medical problem lists. Using the top performing administrative data algorithms, the crude population-level OA prevalence ranged from 11% to 25% and standardized prevalence ranged from 9 to 22% in 2017. Over time, prevalence increased whereas incidence remained stable (~1% annually). Conclusion: Health administrative data have limited sensitivity in adequately identifying all OA patients and appear to be more sensitive at detecting OA patients for whom their physician formally documented their diagnosis in medical problem lists than individuals who have their diagnosis documented outside of problem lists. Irrespective of the algorithm used, OA prevalence has increased over the past decade while annual incidence has been stable.

Improving Care for Patients With or at Risk for Chronic Kidney Disease Using Electronic Medical Record Interventions: A Pragmatic Cluster-Randomized Trial Protocol.

BACKGROUND: Many patients with or at risk for chronic kidney disease (CKD) in the primary care setting are not receiving recommended care. OBJECTIVE: The objective of this study is to determine whether a multifaceted, low-cost intervention compared with usual care improves the care of patients with or at risk for CKD in the primary care setting. DESIGN: A pragmatic cluster-randomized trial, with an embedded qualitative process evaluation, will be conducted. SETTING: The study population comes from the Electronic Medical Record Administrative data Linked Database®, which includes clinical data for more than 140 000 rostered adults cared for by 194 family physicians in 34 clinics across Ontario, Canada. The 34 primary care clinics will be randomized to the intervention or control group. INTERVENTION: The intervention group will receive resources from the "CKD toolkit" to help improve care including practice audit and feedback, printed educational materials for physicians and patients, electronic decision support and reminders, and implementation support. MEASUREMENTS: Patients with or at risk for CKD within participating clinics will be identified using laboratory data in the electronic medical records. Outcomes will be assessed after dissemination of the CKD tools and after 2 rounds of feedback on performance on quality indicators have been sent to the physicians using information from the electronic medical records. The primary outcome is the proportion of patients aged 50 to 80 years with nondialysis-dependent CKD who are on a statin. Secondary outcomes include process of care measures such as screening tests, CKD recognition, monitoring tests, angiotensin-converting enzyme inhibitor or angiotensin receptor blocker prescriptions, blood pressure targets met, and nephrologist referral. Hierarchical analytic modeling will be performed to account for clustering. Semistructured interviews will be conducted with a random purposeful sample of physicians in the intervention group to understand why the intervention achieved the observed effects. CONCLUSIONS: If our intervention improves care, then the CKD toolkit can be adapted and scaled for use in other primary care clinics which use electronic medical records. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT02274298.

CONTEXTE: On observe que de nombreux patients atteints ou à risque de développer de l'insuffisance rénale chronique (IRC) ne reçoivent pas les soins recommandés dans le cadre des soins de première ligne. OBJECTIF: L'étude vise à déterminer si le recours à une intervention multidimensionnelle et moins coûteuse par rapport aux soins habituellement dispensés améliore les soins prodigués aux patients atteints ou susceptibles de développer de l'IRC dans le cadre des soins primaires. MODÈLE D'ÉTUDE: Il s'agit d'un essai pragmatique randomisé par grappes, auquel on a incorporé une évaluation qualitative. CADRE DE L'ÉTUDE: La population étudiée provient de la base de données EMRALD® (Electronic Medical Record Administrative data Linked Database), qui inclut les données cliniques de plus de 140 000 adultes inscrits soignés par 194 médecins de famille répartis dans 34 cliniques partout en Ontario. Les 34 cliniques de soins de santé de première ligne seront randomisées aléatoirement dans le groupe contrôle ou le groupe d'intervention. GROUPE D'INTERVENTION: Les participants du groupe d'intervention recevront des ressources provenant d'une « boîte d'outils IRC ¼ visant à améliorer les soins. Ce guide comprendra notamment un audit de la pratique et de la rétroaction, du matériel didactique imprimé destiné aux médecins et aux patients, des outils électroniques d'aide à la décision, des rappels par voie électronique ainsi que du soutien à la mise en œuvre. MESURES: Les patients atteints ou à risque de développer de l'IRC au sein des cliniques participantes seront sélectionnés à l'aide des données de laboratoire inscrites dans les dossiers médicaux électroniques. Les résultats seront évalués après la distribution des « boîtes d'outils IRC ¼ et deux rondes de rétroaction sur le rendement des indicateurs de qualité qui auront été envoyés aux médecins à l'aide des informations contenues dans les dossiers médicaux électroniques. Le résultat principal attendu sera une différence entre les deux groupes dans la proportion de patients âgés de 50 à 80 ans atteints d'IRC, non dépendants de la dialyse, et sous traitement par une statine. Les résultats secondaires comprendront les processus de mesure des soins tels que les tests de dépistage, la constatation de l'IRC, les tests de contrôle, une ordonnance d'un inhibiteur de l'enzyme de conversion de l'angiotensine ou d'un antagoniste du récepteur de l'angiotensine, la rencontre d'une valeur cible de tension artérielle, et le référencement pour un suivi par un néphrologue. La modélisation analytique hiérarchique sera effectuée en prenant compte de la randomisation. Des entretiens semi-directifs seront menés auprès d'un échantillon aléatoire ciblé de médecins du groupe d'intervention afin de comprendre pourquoi l'intervention a permis d'atteindre les effets observés. CONCLUSIONS: Si notre modèle d'intervention parvient à améliorer les soins, la « boîte d'outils IRC ¼ pourra être adaptée et échelonnée en vue d'une utilisation dans d'autres cliniques de soins de première ligne qui utilisent des dossiers médicaux électroniques. ENREGISTREMENT DES ESSAIS: Identifiant ClinicalTrials.gov: NCT02274298.

Identification of Physician-Diagnosed Alzheimer's Disease and Related Dementias in Population-Based Administrative Data: A Validation Study Using Family Physicians' Electronic Medical Records.

BACKGROUND: Population-based surveillance of Alzheimer's and related dementias (AD-RD) incidence and prevalence is important for chronic disease management and health system capacity planning. Algorithms based on health administrative data have been successfully developed for many chronic conditions. The increasing use of electronic medical records (EMRs) by family physicians (FPs) provides a novel reference standard by which to evaluate these algorithms as FPs are the first point of contact and providers of ongoing medical care for persons with AD-RD. OBJECTIVE: We used FP EMR data as the reference standard to evaluate the accuracy of population-based health administrative data in identifying older adults with AD-RD over time. METHODS: This retrospective chart abstraction study used a random sample of EMRs for 3,404 adults over 65 years of age from 83 community-based FPs in Ontario, Canada. AD-RD patients identified in the EMR were used as the reference standard against which algorithms identifying cases of AD-RD in administrative databases were compared. RESULTS: The highest performing algorithm was "one hospitalization code OR (three physician claims codes at least 30 days apart in a two year period) OR a prescription filled for an AD-RD specific medication" with sensitivity 79.3% (confidence interval (CI) 72.9-85.8%), specificity 99.1% (CI 98.8-99.4%), positive predictive value 80.4% (CI 74.0-86.8%), and negative predictive value 99.0% (CI 98.7-99.4%). This resulted in an age- and sex-adjusted incidence of 18.1 per 1,000 persons and adjusted prevalence of 72.0 per 1,000 persons in 2010/11. CONCLUSION: Algorithms developed from health administrative data are sensitive and specific for identifying older adults with AD-RD.

Wait times to rheumatology care for patients with rheumatic diseases: a data linkage study of primary care electronic medical records and administrative data.

BACKGROUND: The Wait Time Alliance recently established wait time benchmarks for rheumatology consultations in Canada. Our aim was to quantify wait times to primary and rheumatology care for patients with rheumatic diseases. METHODS: We identified patients from primary care practices in the Electronic Medical Record Administrative data Linked Database who had referrals to Ontario rheumatologists over the period 2000-2013. To assess the full care pathway, we identified dates of symptom onset, presentation in primary care and referral from electronic medical records. Dates of rheumatologist consultations were obtained by linking with physician service claims. We determined the duration of each phase of the care pathway (symptom onset to primary care encounter, primary care encounter to referral, and referral to rheumatologist consultation) and compared them with established benchmarks. RESULTS: Among 2430 referrals from 168 family physicians, 2015 patients (82.9%) were seen by 146 rheumatologists within 1 year of referral. Of the 2430 referrals, 2417 (99.5%) occurred between 2005 and 2013. The main reasons for referral were osteoarthritis (32.4%) and systemic inflammatory rheumatic diseases (30.6%). Wait times varied by diagnosis and geographic region. Overall, the median wait time from referral to rheumatologist consultation was 74 (interquartile range 27-101) days; it was 66 (interquartile range 18-84) days for systemic inflammatory rheumatic diseases. Wait time benchmarks were not achieved, even for the most urgent types of referral. For systemic inflammatory rheumatic diseases, most of the delays occurred before referral. INTERPRETATION: Rheumatology wait times exceeded established benchmarks. Targeted efforts are needed to promote more timely access to both primary and rheumatology care. Routine linkage of electronic medical records with administrative data may help fill important gaps in knowledge about waits to primary and specialty care.

The association between diabetes and breast cancer stage at diagnosis: a population-based study.

Women with diabetes have higher breast cancer incidence and mortality. The purpose of this study was to examine the impact of diabetes on stage at breast cancer diagnosis, as a possible reason for their higher mortality. Using population-based health databases from Ontario, Canada, this retrospective cohort study examined stage at diagnosis (II, III, or IV vs I) among women aged 20-105 years who were newly diagnosed with invasive breast cancer between 2007 and 2012. We compared those with diabetes to those without diabetes. Diabetes was defined based on medical records using a validated algorithm. Among 38,407 women with breast cancer, 6115 (15.9 %) women had diabetes. Breast cancer patients with diabetes were significantly more likely to present with advanced-stage breast cancer than those without diabetes. After adjustment for mammograms and other covariates, diabetes was associated with a significantly increased risk of Stage II [adjusted odds ratio (aOR) 1.14, 95 % confidence interval (CI) 1.07, 1.22], Stage III (aOR 1.21, 95 % CI 1.11, 1.33), and Stage IV (aOR 1.16, 95 % CI 1.01, 1.33) versus Stage I breast cancer. Women with diabetes had a higher risk of lymph node metastases (aOR 1.16, 95 % CI 1.06, 1.27) and tumors with size over 2 cm (aOR 1.16, 95 % CI 1.06, 1.28). Diabetes was associated with more advanced-stage breast cancer, even after accounting for differences in screening mammogram use and other factors. Our findings suggest that diabetes may predispose to more aggressive breast cancer, which may be a contributor to their higher cancer mortality.

How infectious disease outbreaks affect community-based primary care physicians: comparing the SARS and H1N1 epidemics.

OBJECTIVE: To compare how the infectious disease outbreaks H1N1 and severe acute respiratory syndrome (SARS) affected community-based GPs and FPs. DESIGN: A mailed survey sent after the H1N1 outbreak compared with the results of similar survey completed after the SARS outbreak. SETTING: Greater Toronto area in Ontario. PARTICIPANTS: A total of 183 randomly selected GPs and FPs who provided office-based care. MAIN OUTCOME MEASURES: The perceptions of GPs and FPs on how serious infectious disease outbreaks affected their clinical work and personal lives; their preparedness for a serious infectious disease outbreak; and the types of information they want to receive and the sources they wanted to receive information from during a serious infectious disease outbreak. The responses from this survey were compared with the responses of GPs and FPs in the greater Toronto area who completed a similar survey in 2003 after the SARS outbreak. RESULTS: After the H1N1 outbreak, GPs and FPs still had substantial concerns about the effects of serious infectious disease outbreaks on the health of their family members. Physicians made changes to various office practices in order to manage and deal with patients with serious infectious diseases. They expressed concerns about the effects of an infectious disease on the provision of health care services. Also, physicians wanted to quickly receive accurate information from the provincial government and their medical associations. CONCLUSION: Serious community-based infectious diseases are a personal concern for GPs and FPs, and have considerable effects on their clinical practice. Further work examining the timely flow of relevant information through different health care sectors and government agencies still needs to be undertaken.

An administrative data validation study of the accuracy of algorithms for identifying rheumatoid arthritis: the influence of the reference standard on algorithm performance.

BACKGROUND: We have previously validated administrative data algorithms to identify patients with rheumatoid arthritis (RA) using rheumatology clinic records as the reference standard. Here we reassessed the accuracy of the algorithms using primary care records as the reference standard. METHODS: We performed a retrospective chart abstraction study using a random sample of 7500 adult patients under the care of 83 family physicians contributing to the Electronic Medical Record Administrative data Linked Database (EMRALD) in Ontario, Canada. Using physician-reported diagnoses as the reference standard, we computed and compared the sensitivity, specificity, and predictive values for over 100 administrative data algorithms for RA case ascertainment. RESULTS: We identified 69 patients with RA for a lifetime RA prevalence of 0.9%. All algorithms had excellent specificity (>97%). However, sensitivity varied (75-90%) among physician billing algorithms. Despite the low prevalence of RA, most algorithms had adequate positive predictive value (PPV; 51-83%). The algorithm of "[1 hospitalization RA diagnosis code] or [3 physician RA diagnosis codes with ≥1 by a specialist over 2 years]" had a sensitivity of 78% (95% CI 69-88), specificity of 100% (95% CI 100-100), PPV of 78% (95% CI 69-88) and NPV of 100% (95% CI 100-100). CONCLUSIONS: Administrative data algorithms for detecting RA patients achieved a high degree of accuracy amongst the general population. However, results varied slightly from our previous report, which can be attributed to differences in the reference standards with respect to disease prevalence, spectrum of disease, and type of comparator group.

Feasibility and validity of the self-administered computerized assessment of mild cognitive impairment with older primary care patients.

We investigated whether a validated computerized cognitive test, the Computerized Assessment of Mild Cognitive Impairment (CAMCI), could be independently completed by older primary care patients. We also determined the optimal cut-off for the CAMCI global risk score for mild cognitive impairment against an independent neuropsychological reference standard. All eligible patients aged 65 years and older, seen consecutively over 2 months by 1 family practice of 13 primary care physicians, were invited to participate. Patients with a diagnosis or previous work-up for dementia were excluded. Primary care physicians indicated whether they, the patient, or family had concerns about each patient's cognition. A total of 130 patients with cognitive concerns and a matched sample of 133 without cognitive concerns were enrolled. The CAMCI was individually administered after instructions to work independently. Comments were recorded verbatim. A total of 259 (98.5%) completed the entire CAMCI. Two hundred and forty-one (91.6%) completed it without any questions or after simple acknowledgment of their question. Lack of computer experience was the only patient characteristic that decreased the odds of independent CAMCI completion. These results support the feasibility of using self-administered computerized cognitive tests with older primary care patients, given the increasing reliance on computers by people of all ages. The optimal cut-off score had a sensitivity of 80% and specificity of 74%.

Access to rheumatologists among patients with newly diagnosed rheumatoid arthritis in a Canadian universal public healthcare system.

OBJECTIVES: Our objective was to estimate the percentage of patients with incident rheumatoid arthritis (RA) who were seen by a rheumatologist within 3, 6 and 12 months of suspected diagnosis by a family physician, and assess what factors may influence the time frame with which patients are seen. SETTING: Ontario, Canada. PARTICIPANTS: Over 2000-2009, we studied patients with incident RA who were initially diagnosed by a family physician. PRIMARY AND SECONDARY OUTCOME MEASURES: We assessed secular trends in rheumatology encounters and differences between patients who received versus did not receive rheumatology care. We performed hierarchical logistic regression analyses to determine whether receipt of rheumatology care was associated with patient, primary care physician and geographical factors. RESULTS: Among 19 760 patients with incident RA, 59%, 75% and 84% of patients were seen by a rheumatologist within 3, 6 and 12 months, respectively. The prevalence of initial consultations within 3 months did not increase over time; however, access within 6 and 12 months increased over time. Factors positively associated with timely consultations included higher regional rheumatology supply (adjusted OR (aOR) 1.35 (95% CI 1.13 to 1.60)) and higher patient socioeconomic status (aOR 1.18 (95% CI 1.07 to 1.30)). Conversely, factors inversely associated with timely consultations included remote patient residence (aOR 0.51 (95% CI 0.41 to 0.64)) and male family physicians (aOR 0.88 (95% CI 0.81 to 0.95)). CONCLUSIONS: Increasing access to rheumatologists within 6 and 12 months occurred over time; however, consultations within 3 months did not change over time. Measures of poor access (such as proximity to and density of rheumatologists) were negatively associated with timely consultations. Additional factors that contributed to disparities in access included patient socioeconomic status and physician sex.

Assessing the validity of using administrative data to identify patients with epilepsy.

OBJECTIVE: Previous validation studies assessing the use of administrative data to identify patients with epilepsy have used targeted sampling or have used a reference standard of patients in the neurologist, hospital, or emergency room setting. Therefore, the validity of using administrative data to identify patients with epilepsy in the general population has not been previously assessed. The purpose of this study was to determine the validity of using administrative data to identify patients with epilepsy in the general population. METHODS: A retrospective chart abstraction study was performed using primary care physician records from 83 physicians distributed throughout Ontario and contributing data to the Electronic Medical Record Administrative data Linked Database (EMRALD) A random sample of 7,500 adult patients, from a possible 73,014 eligible, was manually chart abstracted to identify patients who had ever had epilepsy. These patients were used as a reference standard to test a variety of administrative data algorithms. RESULTS: An algorithm of three physician billing codes (separated by at least 30 days) in 2 years or one hospitalization had a sensitivity of 73.7% (95% confidence interval [CI] 64.8-82.5%), specificity of 99.8% (95% CI 99.6-99.9%), positive predictive value (PPV) of 79.5% (95% CI 71.1-88.0%), and negative predictive value (NPV) of 99.7% (95% CI 99.5-99.8%) for identifying patients who had ever had epilepsy. SIGNIFICANCE: The results of our study showed that administrative data can reasonably accurately identify patients who have ever had epilepsy, allowing for a "lifetime" population prevalence determination of epilepsy in Ontario and the rest of Canada with similar administrative databases. This will facilitate future studies on population level patterns and outcomes of care for patients living with epilepsy.

A qualitative descriptive study on the alignment of care goals between older persons with multi-morbidities, their family physicians and informal caregivers.

BACKGROUND: Goal setting is a recommended approach in clinical care that can help individuals with multi-morbidities and their caregivers manage chronic conditions. In this paper, the types of goals that were important for older persons with multi-morbidities were explored from the perspectives of patients, their caregivers and physicians. Comparisons of goals were made across each patient, caregiver and physician triad to determine alignment. METHODS: The study was a qualitative descriptive study facilitated through semi-structured one-on-one interviews. The study took place between May and October 2012 at a Family Health Team located in Ontario, Canada. The sample included 28 family medicine patients, their informal caregivers and family physicians. Socio-demographic data were analyzed via descriptive statistics in SPSS Version 17. Open ended questions pertaining to patient goals of care were analyzed thematically using NVivo9. Themes were derived on patient care goals for each of the participant groups (patients, caregivers and family physicians). Following this, alignment of goals across each of the triads was examined. Goal alignment was defined as concurrence on at least one goal by all three parties in a particular triad (i.e., patient, caregiver and family physician). RESULTS: Just over half of the patients were male (56%); they had an average age of 82.3 years and 4.61 health conditions. Most of the caregivers were female (82%); and 61% were a spouse of the care recipient. At the aggregate level, common goals expressed among patients, caregivers and family physicians were the maintenance of functional independence of patients and the management of their symptoms or functional challenges. Despite these common goals at the aggregate level, little alignment of goals was found when looking across patient-caregiver and physician triads. Lack of alignment tended to occur when patients had unstable or declining functional or cognitive health; when safety threats were noted; and when enhanced care services were required. CONCLUSIONS: The data suggest that goal divergence tends to occur when patients are less medically stable. While goal divergence may be expected due to the different roles and responsibilities of each of the players involved, these perspectives should be illuminated when building care plans. Further research is required to observe the extent to which goal setting occurs in family practice as well as how it can be embedded as a standard of practice.

The rising burden of rheumatoid arthritis surpasses rheumatology supply in Ontario.

OBJECTIVES: Accurate data on the burden of rheumatoid arthritis (RA) are scarce, but critical in helping health care providers and decision makers to optimize clinical and public health strategies for disease management. We quantified the burden of RA in Ontario from 1996 to 2010 by age, sex and health planning region. METHODS: We used the Ontario Rheumatoid Arthritis administrative Database (ORAD), a validated population-based cohort of all Ontarians with RA, to estimate the crude prevalence and incidence of RA among men and women, and by age group from 1996 to 2010. Burden by area of patient residence and rheumatology supply also were determined. RESULTS: The number of RA patients increased over time, from 42,734 Ontarians (0.5%) in 1996 to 97,499 (0.9%) in 2010. On average 5,830 new RA patients were diagnosed each year. In 2010, the burden was higher among females (1.3%) than males (0.5%) and increased with age, with almost half of all RA patients aged 65 years and older. The burden was higher in northern communities (1.0%) than in southern urban areas (0.7%). During the study period, the number of rheumatologists practicing in Ontario remained unchanged (approximately 160). CONCLUSION: Over a 15-year period, the number of RA patients more than doubled with no concomitant increase in the number of practicing rheumatologists. We observed considerable regional variation in burden, with the highest rates observed in the north. Our findings highlight the need for regional approaches to the planning and delivery of RA care in order to manage the growing burden.

Effects of implementing electronic medical records on primary care billings and payments: a before-after study.

BACKGROUND: Several barriers to the adoption of electronic medical records (EMRs) by family physicians have been discussed, including the costs of implementation, impact on work flow and loss of productivity. We examined billings and payments received before and after implementation of EMRs among primary care physicians in the province of Ontario. We also examined billings and payments before and after switching from a fee-for-service to a capitation payment model, because EMR implementation coincided with primary care reform in the province. METHODS: We used information from the Electronic Medical Record Administrative Data Linked Database (EMRALD) to conduct a retrospective before-after study. The EMRALD database includes EMR data extracted from 183 community-based family physicians in Ontario. We included EMRALD physicians who were eligible to bill the Ontario Health Insurance Plan at least 18 months before and after the date they started using EMRs and had completed a full 18-month period before Mar. 31, 2011, when the study stopped. The main outcome measures were physicians' monthly billings and payments for office visits and total annual payments received from all government sources. Two index dates were examined: the date physicians started using EMRs and were in a stable payment model (n = 64) and the date physicians switched from a fee-for-service to a capitation payment model (n = 42). RESULTS: Monthly billings and payments for office visits did not decrease after the implementation of EMRs. The overall weighted mean annual payment from all government sources increased by 27.7% after the start of EMRs among EMRALD physicians; an increase was also observed among all other primary care physicians in Ontario, but it was not as great (14.4%). There was a decline in monthly billings and payments for office visits after physicians changed payment models, but an increase in their overall annual government payments. INTERPRETATION: Implementation of EMRs by primary care physicians did not result in decreased billings or government payments for office visits. Further economic analyses are needed to measure the effects of EMR implementation on productivity and the costs of implementing an EMR system, including the costs of nonclinical work by physicians and their staff.

Office-based procedures among urban and rural family physicians in Ontario.

OBJECTIVE: To compare FP and GP performance of office-based procedures between urban and rural practices. DESIGN: Descriptive cohort study using health administrative data. SETTING: Ontario. PARTICIPANTS: All FPs and GPs who billed the Ontario Health Insurance Plan for at least 1 office-based procedure between January 1 and December 31, 2006 (N = 8648). MAIN OUTCOME MEASURES: Ontario Health Insurance Plan billings for office-based procedures were adjusted by full-time equivalent (FTE) so that the means are for 1 FTE. Office-based procedures were grouped into 1) surgical procedures, 2) injections and immunizations, 3) electrocardiograms (ECGs), and 4) venipunctures and laboratory tests. The analyses were stratified for FP and GP age, sex, rurality of practice, and participation in a primary care model. RESULTS: There were no substantial differences between FPs and GPs in rural practices compared with those in more urban practices with respect to surgical procedures. Rural FPs and GPs had lower mean numbers of injections and immunizations, ECGs, and venipunctures and laboratory tests than FPs and GPs practising in urban areas. Family physicians and GPs in primary care models had a lower mean number of surgical procedures but a higher adjusted mean number of injections and immunizations, ECGs, and venipunctures and laboratory tests. CONCLUSION: For those procedures that are not dependent on specialist backup or access to more advanced technology, there were no substantial differences between rural and urban FPs and GPs. All comprehensive FPs and GPs should be able to provide these services to their patients. Training programs for all family medicine residents should ensure future FPs and GPs are able to perform these procedures.

Assessing methods for measurement of clinical outcomes and quality of care in primary care practices.

PURPOSE: To evaluate the appropriateness of potential data sources for the population of performance indicators for primary care (PC) practices. METHODS: This project was a cross sectional study of 7 multidisciplinary primary care teams in Ontario, Canada. Practices were recruited and 5-7 physicians per practice agreed to participate in the study. Patients of participating physicians (20-30) were recruited sequentially as they presented to attend a visit. Data collection included patient, provider and practice surveys, chart abstraction and linkage to administrative data sets. Matched pairs analysis was used to examine the differences in the observed results for each indicator obtained using multiple data sources. RESULTS: Seven teams, 41 physicians, 94 associated staff and 998 patients were recruited. The survey response rate was 81% for patients, 93% for physicians and 83% for associated staff. Chart audits were successfully completed on all but 1 patient and linkage to administrative data was successful for all subjects. There were significant differences noted between the data collection methods for many measures. No single method of data collection was best for all outcomes. For most measures of technical quality of care chart audit was the most accurate method of data collection. Patient surveys were more accurate for immunizations, chronic disease advice/information dispensed, some general health promotion items and possibly for medication use. Administrative data appears useful for indicators including chronic disease diagnosis and osteoporosis/ breast screening. CONCLUSIONS: Multiple data collection methods are required for a comprehensive assessment of performance in primary care practices. The choice of which methods are best for any one particular study or quality improvement initiative requires careful consideration of the biases that each method might introduce into the results. In this study, both patients and providers were willing to participate in and consent to, the collection and linkage of information from multiple sources that would be required for such assessments.