Racial and socioeconomic disparities in the outcome of patients hospitalized for acute myocardial infarction (AMI), heart failure (HF), and pneumonia are attributable not to disparities in care within each hospital, but rather to broader systemic factors, new research suggests.
Investigators analyzed 12 cohorts of hospitals — each comprised of 400 to more than 1200 hospitals and 145,000 to more than 700,000 patients — to see if mortality and readmission rates for MI, HF, and pneumonia were affected by race or socioeconomic factors.
In contrast, there were minimal differences associated with socioeconomic status.
“We found major differences in race. Interestingly, that black patients tended to have lower mortality rates for heart attack, heart failure, and pneumonia, but higher readmission rates consistently across all hospitals,” Harlan M. Krumholz, MD, SM, professor of medicine and professor in the Institute for Social and Policy studies of Investigative Medicine and Public Health, and director of the Center for Outcomes Research and Evaluation, Yale–New Haven Hospital in Connecticut, told theheart.org | Medscape Cardiology.
“Good hospitals for white patients tended to be equally good for black patients, suggesting systemic effects in our healthcare system that are affecting all hospitals, not just a specific pocket of hospitals in a specific area,” he said.
It is, however, “important to know whether hospital performance differs according to patients’ race and socioeconomic status” (SES).
“We were interested in determining whether there were between-hospital or within-hospital differences in outcomes of patient by income and race, and whether top hospitals for white or upper-class patients were also top hospitals for black or lower-income patients,” Krumholz said.
To investigate whether there were within-hospital differences in performance between patient groups defined by race or SES, the researchers calculated 30-day risk-standardized outcome measures of mortality rates (RSMRs) and readmission rates (RSRRs) by race and neighborhood subgroups at each hospital.
The calculations controlled for age, sex, and clinical covariates by race and neighborhood subgroups at each hospital.
The researchers then compared the corresponding ratios using intraclass correlation coefficients.
A significant proportion of hospitals (between 74% and 91%) lacked sufficient racial and socioeconomic diversity to be included in the analysis, and the number of hospitals eligible for analysis varied among cohorts.
The researchers were able to construct 12 analysis cohorts, ranging in size from 418 hospitals (in the analysis of AMI mortality by race) to 1265 hospitals (in the analysis of HF readmission by neighborhood income).
Hospitals included in the analysis accounted for a disproportionate number of admissions, with the number of hospital admissions ranging from 144,417 in the analysis of AMI mortality by race to 703,324 in the analysis of HF readmissions by race.
Analyses were performed for each of the 12 analysis cohorts, reflecting the unique combinations of outcomes (mortality and readmission), demographics (race and neighborhood income), and conditions (AMI, HF, and pneumonia).
A smaller proportion of black than of white patients were male (40.8% to 43.9% vs 46.1% to 53.4%).
Black patients tended to be younger than white patients (mean age of each analysis cohort, 77.0 to 78.4 vs 78.2 to 81.3 years).
The proportion of patients from the higher- and lower-income neighborhoods was approximately equal in all the income cohorts.
A smaller proportion of patients from lower- than from higher-income neighborhoods were male (45.4% to 51.8% vs 46.5% to 53.4%).
Additionally, patients from lower-income neighborhoods tended to be slightly younger than those from higher-income neighborhoods (mean age for each analysis cohort, 77.6 to 79.8 vs 78.8 to 81.8 years).
The RSMRs were lower among black patients than white patients (mean [SD] difference for AMI, −0.57 [1.1], P = 0.47; for HF, −4.7 [1.3], P < .001; and for pneumonia, −1.0 [2.0], P = .05).
In contrast, RSRRs were higher among black patients than white patients (mean [SD] difference for AMI, 4.3 [1.4] P < .001; for HF, 2.8 [1.8], P < .001, and for pneumonia, 3.7 [1.3], P < .001).
Intraclass correlation coefficients ranged from 0.68 to 0.79, “indicating that hospitals generally delivered consistent quality to patients of differing races,” the authors comment.
The coefficients in the neighborhood income analysis were slightly lower (0.46 to 0.60), suggesting some heterogeneity in within-hospital performance. However, differences in mortality rates and readmission rates between the two neighborhood income groups were small.
The researchers found no strong, consistent associations between risk-standardized outcomes for white or higher-income neighborhood patients and hospitals‘ proportion of black or lower-income neighborhood patients.
“The higher readmission rates we observed are not attributable to a particular set of hospitals providing especially bad care for black patients because our findings were consistent across all hospitals, showing that these differences tend to occur across the spectrum and are systemic,” Krumholz said.
He acknowledged that it is unclear why black patients had lower mortality rates and higher readmission rates.
“It is possible that there is something about the care they are receiving after discharge that might be affecting the increased need for readmission,” he suggested.
He added, “This finding should get us to redouble our efforts and investigate the underlying mechanism behind our observation.”
Focus on the Larger Picture
This is “an important contribution to the literature on racial disparities in healthcare, especially in cardiac care,” said Quinn Capers IV, MD, associate professor of medicine and program director, interventional cardiology fellowship, Division of Cardiovascular Medicine, Ohio State University College of Medicine, Columbus, when asked for comment by theheart.org | Medscape Cardiology.
The findings “open up the possibility that ‘extra-hospital‘ factors that impact mortality and readmission rates may impact black and white patients differently,” he said.
The lower mortality rate in black patients might be partially explained by the fact that they were significantly younger than the white patients, and the greater likelihood of readmission implicates disparities in follow-up care outside the hospital, Capers explained.
He noted that the patients in the study were all 65 years or older, “so these findings cannot be generalized to younger patients.” Additionally, cohort sizes were small, and nearly three-quarters of American hospitals were excluded because they did not have sufficient diversity in their patient population.
Nevertheless, the study suggests addressing “system-wide” inequities, such as physician-to-patient ratio in geographic areas, systemic racism and the psychological burden it may impose, “while simultaneously addressing local problems, such as hospitals participating in quality-improvement programs and implicit bias training of physicians at individual hospitals,” Capers stated.
“The systemic issues may be societal and may not only affect healthcare, so we need to focus on that larger picture,” Krumholz added.
Krumholz is under contract with the Centers for Medicare & Medicaid Services to develop and maintain performance measures that are publicly reported. He has received research grants through Yale, from Medtronic, and from the US Food and Drug Administration to develop methods for postmarket surveillance of medical devices, and is a recipient of a grant from Medtronic and Johnson & Johnson, through Yale, to develop methods of clinical trial data sharing. He chairs a Cardiac Scientific Advisory Board for UnitedHealth; is a participant/participant representative of the IBM Watson Health Life Sciences Board; is a member of the Advisory Board for Element Science and the Physician Advisory Board for Aetna; and is the founder of Hugo, a personal health information platform. The other authors’ disclosures and sources of funding are listed on the original paper. Capers reports no conflicting interests.
JAMA Network Open. Published online September 7, 2018. Full text