Linda R. Donovan

Comparison of coding of heart failure and comorbidities in administrative and clinical data for use in outcomes research

Background:Despite the potential usefulness of administrative databases for evaluating outcomes, coding of heart failure and associated comorbidities have not been definitively compared with clinical data. Objective:To compare the predictive value of heart failure diagnoses and secondary conditions identified in a large administrative database with chart-based records. Methods:The authors studied 1808 patient records sampled from 14 acute care hospitals and compared clinically recorded data with administrative records from the Canadian Institute for Health Information. The impact of comorbidity coding in the administrative data set according to the Charlson classification was examined in models of 30-day mortality. Results:The positive predictive value (PPV) of a primary diagnosis ICD-9 428 was 94.3% using the Framingham criteria and 88.6% using criteria previously validated with pulmonary capillary wedge pressure. There was reduced prevalence of secondary comorbid conditions in administrative data in comparison with clinical chart data. The specificities and PPV/negative predictive values of administratively identified index comorbidities were high. The sensitivities of index comorbidities were low, but were enhanced by examination of hospitalizations within 1 year prior to the index heart failure admission. Using information from prior hospitalizations modestly enhanced 30-day mortality model performance; however, the odds ratio point estimates of the index and enhanced administrative data sets were consistent with the clinical model. Conclusion:The ICD-9 428 primary diagnosis is highly predictive of heart failure using clinical criteria. Examination of hospitalization data up to 1 year prior to the index admission improves comorbidity detection and may provide enhancements to future studies of heart failure mortality.

Effectiveness of Public Report Cards for Improving the Quality of Cardiac Care The EFFECT Study: A Randomized Trial

CONTEXT Publicly released report cards on hospital performance are increasingly common, but whether they are an effective method for improving quality of care remains uncertain. OBJECTIVE To evaluate whether the public release of data on cardiac quality indicators effectively stimulates hospitals to undertake quality improvement activities that improve health care processes and patient outcomes. DESIGN, SETTING, AND PATIENTS Population-based cluster randomized trial (Enhanced Feedback for Effective Cardiac Treatment [EFFECT]) of 86 hospital corporations in Ontario, Canada, with patients admitted for acute myocardial infarction (AMI) or congestive heart failure (CHF). INTERVENTION Participating hospital corporations were randomized to early (January 2004) or delayed (September 2005) feedback of a public report card on their baseline performance (between April 1999 and March 2001) on a set of 12 process-of-care indicators for AMI and 6 for CHF. Follow-up performance data (between April 2004 and March 2005) also were collected. MAIN OUTCOME MEASURES The coprimary outcomes were composite AMI and CHF indicators based on 12 AMI and 6 CHF process-of-care indicators. Secondary outcomes were the individual process-of-care indicators, a hospital report card impact survey, and all-cause AMI and CHF mortality. RESULTS The publication of the early feedback hospital report card did not result in a significant systemwide improvement in the early feedback group in either the composite AMI process-of-care indicator (absolute change, 1.5%; 95% confidence interval [CI], -2.2% to 5.1%; P = .43) or the composite CHF process-of-care indicator (absolute change, 0.6%; 95% CI, -4.5% to 5.7%; P = .81). During the follow-up period, the mean 30-day AMI mortality rates were 2.5% lower (95% CI, 0.1% to 4.9%; P = .045) in the early feedback group compared with the delayed feedback group. The hospital mortality rates for CHF were not significantly different. CONCLUSION Public release of hospital-specific quality indicators did not significantly improve composite process-of-care indicators for AMI or CHF. TRIAL REGISTRATION http://clinicaltrials.gov Identifier: NCT00187460.

Indicators of quality of care for patients with acute myocardial infarction

Background: There is a wide practice gap between optimal and actual care for patients with acute myocardial infarction in hospitals around the world. We undertook this initiative to develop an updated set of evidence-based indicators to measure and improve the quality of care for this patient population. Methods: A 12-member expert panel was convened in 2007 to develop an updated set of quality indicators for acute myocardial infarction. The panel identified a list of potential indicators after reviewing the scientific literature, clinical practice guidelines and other published quality indicators. To develop the new list of indicators, the panel rated each potential indicator on 4 dimensions (reliability, validity, feasibility and usefulness in improving patient outcomes) and discussed the top-ranked quality indicators at a consensus meeting. Results: Consensus was reached on 38 quality indicators: 17 that would be measurable using chart-abstracted data and 21 that would be measurable using administrative data. Of the 17 chart-review indicators, 13 address pharmacologic and nonpharmacologic care delivered to patients in hospital. In-hospital mortality was recommended as a key outcome indicator. Three system indicators were recommended to measure the collaborative responsiveness of the health care system from the call for help to intervention. It was recommended that hospitals strive for a minimum target benchmark of 90% or greater on process-of-care indicators. Interpretation: Implementation of strategies by clinicians and hospitals to meet target benchmarks on these quality indicators could save the lives of many individuals with acute myocardial infarction.

Validity of the GRACE (Global Registry of Acute Coronary Events) acute coronary syndrome prediction model for six month post-discharge death in an independent data set

Objective: To determine the validity of the GRACE (Global Registry of Acute Coronary Events) prediction model for death six months after discharge in all forms of acute coronary syndrome in an independent dataset of a community based cohort of patients with acute myocardial infarction (AMI). Design: Independent validation study based on clinical data collected retrospectively for a clinical trial in a community based population and record linkage to administrative databases. Setting: Study conducted among patients from the EFFECT (enhanced feedback for effective cardiac treatment) study from Ontario, Canada. Patients: Randomly selected men and women hospitalised for AMI between 1999 and 2001. Main outcome measure: Discriminatory capacity and calibration of the GRACE prediction model for death within six months of hospital discharge in the contemporaneous EFFECT AMI study population. Results: Post-discharge crude mortality at six months for the EFFECT study patients with AMI was 7.0%. The discriminatory capacity of the GRACE model was good overall (C statistic 0.80) and for patients with ST segment elevation AMI (STEMI) (0.81) and non-STEMI (0.78). Observed and predicted deaths corresponded well in each stratum of risk at six months, although the risk was underestimated by up to 30% in the higher range of scores among patients with non-STEMI. Conclusions: In an independent validation the GRACE risk model had good discriminatory capacity for predicting post-discharge death at six months and was generally well calibrated, suggesting that it is suitable for clinical use in general populations.

The Cardiovascular Health in Ambulatory Care Research Team (CANHEART) Using Big Data to Measure and Improve Cardiovascular Health and Healthcare Services

Background—The CArdiovascular HEalth in Ambulatory care Research Team (CANHEART) is conducting a unique, population-based observational research initiative aimed at measuring and improving cardiovascular health and the quality of ambulatory cardiovascular care provided in Ontario, Canada. A particular focus will be on identifying opportunities to improve the primary and secondary prevention of cardiovascular events in Ontario’s diverse multiethnic population. Methods and Results—A population-based cohort comprising 9.8 million Ontario adults ≥20 years in 2008 was assembled by linking multiple electronic survey, health administrative, clinical, laboratory, drug, and electronic medical record databases using encoded personal identifiers. The cohort includes ≈9.4 million primary prevention patients and ≈400 000 secondary prevention patients. Follow-up on clinical events is achieved through record linkage to comprehensive hospitalization, emergency department, and vital statistics administrative databases. Profiles of cardiovascular health and preventive care will be developed at the health region level, and the cohort will be used to study the causes of regional variation in the incidence of major cardiovascular events and other important research questions. Conclusions—Linkage of multiple databases will enable the CANHEART study cohort to serve as a powerful big data resource for scientific research aimed at improving cardiovascular health and health services delivery. Study findings will be shared with clinicians, policy makers, and the public to facilitate population health interventions and quality improvement initiatives.

Validation of the Thrombolysis In Myocardial Infarction (TIMI) risk index for predicting early mortality in a population-based cohort of STEMI and non-STEMI patients

BACKGROUND The Thrombolysis In Myocardial Infarction (TIMI) risk index for the prediction of 30-day mortality was developed and validated in patients with ST-segment elevation myocardial infarction (STEMI) who were being treated with thrombolytics in randomized clinical trials. When tested in clinical registries of patients with STEMI, the index performed poorly in an older (65 years and older) Medicare population, but it was a good predictor of early death among the more representative population on the National Registry of Myocardial Infarction-3 and -4 databases. It has not been tested in a population outside the United States or among non-STEMI patients. METHODS The TIMI risk index was applied to the Enhanced Feedback for Effective Cardiac Treatment (EFFECT) study cohort of 11,510 acute MI patients from Ontario. The models discriminatory capacity and calibration were tested in all patients and in subgroups determined by age, sex, diagnosis and reperfusion status. RESULTS The TIMI risk index was strongly associated with 30-day mortality for both STEMI and non-STEMI patients. The C statistic was 0.82 for STEMI and 0.80 for non-STEMI patients, with overlapping 95% CI. The discriminatory capacity was somewhat lower for patients older than 65 years of age (0.74). The model was well calibrated. CONCLUSIONS The TIMI risk index is a simple, valid and moderately accurate tool for the stratification of risk for early death in STEMI and non-STEMI patients in the community setting. Its routine clinical use is warranted.

The Incidence of Major Cardiovascular Events in Immigrants to Ontario, Canada: The CANHEART Immigrant Study.

Background —Immigrants from ethnic minority groups represent an increasing proportion of the population in many high-income countries but little is known about the causes and amount of variation between various immigrant groups in the incidence of major cardiovascular events. Methods and Results —We conducted the Cardiovascular Health in Ambulatory Care Research Team (CANHEART) Immigrant study, a big data initiative, linking information from Citizenship and Immigration Canadas Permanent Resident database to nine population-based health databases. A cohort of 824 662 first-generation immigrants aged 30 to 74 as of January 2002 from eight major ethnic groups and 201 countries of birth who immigrated to Ontario, Canada between 1985 and 2000 were compared to a reference group of 5.2 million long-term residents. The overall 10-year age-standardized incidence of major cardiovascular events was 30% lower among immigrants compared with long-term residents. East Asian immigrants (predominantly ethnic Chinese) had the lowest incidence overall (2.4 in males, 1.1 in females per 1000 person-years) but this increased with greater duration of stay in Canada. South Asian immigrants, including those born in Guyana had the highest event rates (8.9 in males, 3.6 in females per 1000 person-years), along with immigrants born in Iraq and Afghanistan. Adjustment for traditional risk factors reduced but did not eliminate differences in cardiovascular risk between various ethnic groups and long-term residents. Conclusions —Striking differences in the incidence of cardiovascular events exist among immigrants to Canada from different ethnic backgrounds. Traditional risk factors explain part but not all of these differences.Background— Immigrants from ethnic minority groups represent an increasing proportion of the population in many high-income countries, but little is known about the causes and amount of variation between various immigrant groups in the incidence of major cardiovascular events. Methods and Results— We conducted the Cardiovascular Health in Ambulatory Care Research Team (CANHEART) Immigrant Study, a big data initiative, linking information from Citizenship and Immigration Canada’s Permanent Resident database to 9 population-based health databases. A cohort of 824 662 first-generation immigrants aged 30 to 74 as of January 2002 from 8 major ethnic groups and 201 countries of birth who immigrated to Ontario, Canada between 1985 and 2000 were compared with a reference group of 5.2 million long-term residents. The overall 10-year age-standardized incidence of major cardiovascular events was 30% lower among immigrants than among long-term residents. East Asian immigrants (predominantly ethnic Chinese) had the lowest incidence overall (2.4 in males, 1.1 in females per 1000 person-years), but this increased with greater duration of stay in Canada. South Asian immigrants, including those born in Guyana, had the highest event rates (8.9 in males, 3.6 in females per 1000 person-years), along with immigrants born in Iraq and Afghanistan. Adjustment for traditional risk factors reduced but did not eliminate the differences in cardiovascular risk between various ethnic groups and long-term residents. Conclusions— Striking differences in the incidence of cardiovascular events exist among immigrants to Canada from different ethnic backgrounds. Traditional risk factors explain a part but not all of these differences.

The Risk of Ischemic Heart Disease and Stroke Among Immigrant Populations: A Systematic Review

BACKGROUND The increasing frequency of global migration to Canada and other high-income countries has highlighted the need for information on the risk of ischemic heart disease (IHD) and stroke among migrant populations. METHODS Using the MEDLINE and EMBASE databases, we conducted an English-language literature review of articles published from 2000 to 2014 to study patterns in the incidence of IHD or stroke in migrant populations to high-income countries. Our search revealed 17 articles of interest. All studies stratified immigrants according to country or region of birth, except 2 from Canada and 1 from Denmark, in which all immigrant groups were analyzed together. RESULTS The risk of IHD or stroke varied by country of origin, country of destination, and duration of residence. In our review we found that most migrant groups to Western Europe were at a similar or higher risk of IHD and stroke compared with the host population. Those at a higher risk included many Eastern European, Middle-Eastern, and South Asian immigrants. When duration of residence was considered, it appeared that in most migrants the risk of IHD worsened over time. In contrast, immigrants overall were at lower risk of myocardial infarction and stroke in Ontario compared with long-term residents of Canada. CONCLUSIONS The risks of IHD and stroke vary widely in immigrant populations in Western Europe. Detailed studies of immigrants to Canada according to country of birth and duration of residence should be undertaken to guide future cardiovascular health promotion initiatives.

The Influence of Patient Goals of Care on Performance Measures in Patients Hospitalized for Heart Failure: An Analysis of the EFFECT Registry

Background—Pay for performance programs compare metrics that are risk-adjusted, but goals of care are not considered in current models. We conducted this study to explore the associations between do not resuscitate (DNR) designations, quality of care, and outcomes. Methods and Results—Retrospective cohort study with chart review for inpatient quality metrics, 30 day mortality, and readmissions or death within 30 days of discharge in 96 Ontario hospitals participating in the Enhanced Feedback For Effective Cardiac Treatment (EFFECT) study in 2004/05. Of 8339 patients (mean age 77 years) with new heart failure, 1220 (15%) had DNR documented at admission (admission DNR, varying from 0% to 36% between hospitals) and 892 (11%) were switched from full resuscitation to DNR during their index hospitalization (later DNR). Death at 30 days was more common in patients with admission DNR (27%) or later DNR (35%) than full resuscitation (3%)—admission DNR was a stronger predictor (adjusted OR 8.6, 95% confidence interval 6.8–10.7) than any of the variables currently included in heart failure 30 day mortality risk models. Hospital-level rankings differed considerably if DNR patients were excluded: 22 of the 39 EFFECT hospitals in the top and bottom quintiles for 30 day mortality rates (the usual thresholds for rewards/penalties in current performance-based reimbursement schemes) would not have been in those same quintiles if admission DNR patients were excluded. Conclusions—Alternate goals of care are frequent and important confounders in heart failure comparative studies. Philosophy of care discussions should be considered for inclusion as a potential quality of care indicator.Background— Pay for performance programs compare metrics that are risk-adjusted, but goals of care are not considered in current models. We conducted this study to explore the associations between do not resuscitate (DNR) designations, quality of care, and outcomes. Methods and Results— Retrospective cohort study with chart review for inpatient quality metrics, 30 day mortality, and readmissions or death within 30 days of discharge in 96 Ontario hospitals participating in the Enhanced Feedback For Effective Cardiac Treatment (EFFECT) study in 2004/05. Of 8339 patients (mean age 77 years) with new heart failure, 1220 (15%) had DNR documented at admission (admission DNR, varying from 0% to 36% between hospitals) and 892 (11%) were switched from full resuscitation to DNR during their index hospitalization (later DNR). Death at 30 days was more common in patients with admission DNR (27%) or later DNR (35%) than full resuscitation (3%)—admission DNR was a stronger predictor (adjusted OR 8.6, 95% confidence interval 6.8–10.7) than any of the variables currently included in heart failure 30 day mortality risk models. Hospital-level rankings differed considerably if DNR patients were excluded: 22 of the 39 EFFECT hospitals in the top and bottom quintiles for 30 day mortality rates (the usual thresholds for rewards/penalties in current performance-based reimbursement schemes) would not have been in those same quintiles if admission DNR patients were excluded. Conclusions— Alternate goals of care are frequent and important confounders in heart failure comparative studies. Philosophy of care discussions should be considered for inclusion as a potential quality of care indicator.

Quality of Diabetes and Hyperlipidemia Screening Before a First Myocardial Infarction

BACKGROUND There has been significant attention to the quality of care for acute myocardial infarction (MI). However, little is known about the quality of preventive care before a patients first MI. METHODS We conducted a retrospective, cohort analysis of 5688 patients admitted with their first MI to 96 acute care hospitals in Ontario, Canada, from April 2004 to March 2005 using the Enhanced Feedback For Effective Cardiac Treatment clinical study database. We calculated rates of screening for diabetes and hyperlipidemia according to guidelines using linkages to the Ontario Health Insurance Plan database. Screening rates were stratified by age, sex, socioeconomic status, and number of primary care visits in the past 5 years. RESULTS Among the 5688 eligible patients, 27.1% did not receive serum cholesterol screening in the 5 years preceding their MI and 27.5% of patients did not receive a fasting blood glucose or glucose tolerance test in the 3 years before their MI. Women were more likely to be screened than men. Screening rates generally increased with age and were similar across socioeconomic categories. There was a positive association between the number of primary care visits and the likelihood of being screened. CONCLUSIONS A significant number of patients admitted with their first MI were not screened for important modifiable risk factors. Opportunities for the prevention of coronary disease are being missed. More emphasis is needed on identifying risk factors before the development of acute coronary disease.