The influence of patient race on the use of diagnostic imaging in United States emergency departments: data from the National Hospital Ambulatory Medical Care survey

Racial inequities in point-of-care ultrasound for pregnancy

STUDY OBJECTIVE

Racial inequities are pervasive throughout healthcare. We sought to assess if race and ethnicity are associated with emergency department (ED) point-of-care ultrasound (POCUS) usage compared with radiology-ordered ultrasounds as our primary outcome and a secondary outcome of nurse-driven ultrasound ordering for early pregnancy.

METHODS

In this retrospective, observational cohort study between June 2015 and December 2021, we assessed ED physician POCUS use in relation to Radiology (RADUS) ultrasound for first trimester pregnancy with race and ethnicity as our primary variable. A secondary outcome assessed if race and ethnicity impacted nursing-driven ultrasound ordering. Univariate and multivariate logistic regression models were created to test relationships and interactions with clinical variables. Given the overlap of language and race/ethnicity, a multivariate model with language as the primary variable was included.

RESULTS

No significant differences based on race and ethnicity were found for the selection of POCUS versus RADUS (n = 2337: χ2 = 5.25, p = 0.155). For the secondary outcome, 1694 of 7662 (22.1 %) patients received a nurse ultrasound order. Hispanic/Latino patients had increased odds of receiving a nurse-driven order (aOR 1.25, 95 % CI 1.009-1.549) and those of other or unknown race/ethnicity (aOR 1.357, 95 %CI 1.043-1.765) when language was excluded; in addition to Non-English speakers (OR 1.213, 95 %CI 1.072-1.372) with race excluded.

CONCLUSIONS

For first trimester pregnancy complaints, race and ethnicity did not alter POCUS usage by ED physicians. Secondary analysis showed race and ethnicity differences in nurse-driven orders, however collinearity between the primary outcome and language makes it difficult to assess the magnitude of these factors.

Radiology Utilization in an Academic Center Partnership With a Federally Qualified Health Center: A Cross-Sectional Study

OBJECTIVE

Federally qualified health centers (FQHCs) serve medically underserved populations. In 2013, UW Health, the health system of the University of Wisconsin, partnered with Access Community Health Centers (ACHC), the FQHC network in Madison, Wisconsin, to provide on-site outpatient imaging. This study characterized radiography utilization associated with the UW Health-ACHC partnership compared with other UW Health outpatient imaging sites.

METHODS

We included health record data from January 2013 to December 2022 on all outpatient radiographs completed at UW Health sites. We compared characteristics between patients ever seen at ACHC clinics with patients seen only at non-ACHC UW Health clinics using χ2 and t tests. Logistic regression was used to assess factors associated with imaging utilization at ACHC.

RESULTS

Over the study period, 4% (23,794 of 650,685) of imaging encounters occurred at ACHC and 4% (10,986 of 246,104) of patients used ACHC facilities at least once. ACHC clinic patients were younger (41 versus 42) and more often female (55% versus 53%), Black or African American (22% versus 5%), Hispanic or Latino (34% versus 4%), with Medicaid (33% versus 9%), uninsured (18% versus 4%), and living in metropolitan areas (98% versus 88%) with higher Social Deprivation Index scores (53 versus 31) (P < .001). In multivariable analyses, patients from racial or ethnic minority groups, without commercial insurance, residing in a metropolitan area, and with a non-English primary language were more likely to ever use ACHC radiography services (P < .001).

DISCUSSION

FQHCs represent trusted, community health centers serving medically underserved populations. Partnerships between academic institutions and FQHCs can increase geographic imaging access among these groups.

Trends in computed tomography utilization among emergency department patients with foot and ankle trauma

The aim of this study was to assess the number of foot/ankle computed tomography (CT) exams ordered per encounter for patients presenting to the emergency department (ED) with foot and ankle trauma over a 5-year period. Secondary aims included evaluating the positivity rate of foot/ankle CT exams and identifying factors associated with receiving a CT foot/ankle. This retrospective study analyzed data from a large urban Level-1 trauma center between 2016 and 2021. Patients were identified by charted chief complaints related to foot and ankle trauma. The primary outcome was the number of CT foot/ankle exams ordered per patient in a given period. A univariate chi-square analysis was conducted to evaluate differences in patient presentations and imaging rates across the study period. Over the 5-year span, there were 9,845 patient encounters, with a significant increase in CT foot/ankle orders from 2.4 % to 6.6 % (p < 0.001). The CT positivity rate, defined as CTs with positive findings, declined from 95.2 % in 2016 to 84.1 % in 2021 (p < 0.001). Black patients had lower odds of receiving CT scans compared to White patients, as did Medicare recipients compared to Medicaid recipients (p < 0.001). Factors such as age (OR: 1.02 per year), year of visit (OR: 2.66 for 2021), time of day (OR: 1.62 for evening arrivals), and arrival by EMS (OR: 5.60) were significantly associated with higher CT order rates. This study highlights a marked increase in CT utilization for foot and ankle trauma with a corresponding decline in the rate of positive findings. Further research is necessary to explore the reasons behind this trend and to identify potential workflow or protocol adjustments to improve imaging efficacy.

Demographic bias of expert-level vision-language foundation models in medical imaging

Advances in artificial intelligence (AI) have achieved expert-level performance in medical imaging applications. Notably, self-supervised vision-language foundation models can detect a broad spectrum of pathologies without relying on explicit training annotations. However, it is crucial to ensure that these AI models do not mirror or amplify human biases, disadvantaging historically marginalized groups such as females or Black patients. In this study, we investigate the algorithmic fairness of state-of-the-art vision-language foundation models in chest x-ray diagnosis across five globally sourced datasets. Our findings reveal that compared to board-certified radiologists, these foundation models consistently underdiagnose marginalized groups, with even higher rates seen in intersectional subgroups such as Black female patients. Such biases present over a wide range of pathologies and demographic attributes. Further analysis of the model embedding uncovers its substantial encoding of demographic information. Deploying medical AI systems with biases can intensify preexisting care disparities, posing potential challenges to equitable healthcare access and raising ethical questions about their clinical applications.

Developing and Applying the BE-FAIR Equity Framework to a Population Health Predictive Model: A Retrospective Observational Cohort Study

BACKGROUND

Population health programs rely on healthcare predictive models to allocate resources, yet models can perpetuate biases that exacerbate health disparities among marginalized communities.

OBJECTIVE

We developed the Bias-reduction and Equity Framework for Assessing, Implementing, and Redesigning (BE-FAIR) healthcare predictive models, an applied framework tested within a large health system using a population health predictive model, aiming to minimize bias and enhance equity.

DESIGN

Retrospective cohort study conducted at an academic medical center. Data collected from September 30, 2020, to October 1, 2022, were analyzed to assess bias resulting from model use.

PARTICIPANTS

Primary care or payer-attributed patients at the medical center identified through electronic health records and claims data. Participants were stratified by race-ethnicity, gender, and social vulnerability defined by the Healthy Places Index (HPI).

INTERVENTION

BE-FAIR implementation involved steps such as an anti-racism lens application, de-siloed team structure, historical intervention review, disaggregated data analysis, and calibration evaluation.

MAIN MEASURES

The primary outcome was the calibration and discrimination of the model across different demographic groups, measured by logistic regression and area under the receiver operating characteristic curve (AUROC).

RESULTS

The study population consisted of 114,311 individuals with a mean age of 43.4 years (SD 24.0 years), 55.4% female, and 59.5% white/Caucasian. Calibration differed by race-ethnicity and HPI with significantly lower predicted probabilities of hospitalization for African Americans (0.129±0.051, p=0.016), Hispanics (0.133±0.047, p=0.004), AAPI (0.120±0.051, p=0.018), and multi-race (0.245±0.087, p=0.005) relative to white/Caucasians and for individuals in low HPI areas (0 - 25%, 0.178±0.042, p<0.001; 25 - 50%, 0.129±0.044, p=0.003). AUROC values varied among demographic groups.

CONCLUSIONS

The BE-FAIR framework offers a practical approach to address bias in healthcare predictive models, guiding model development, and implementation. By identifying and mitigating biases, BE-FAIR enhances the fairness and equity of healthcare delivery, particularly for minoritized groups, paving the way for more inclusive and effective population health strategies.

Subarachnoid haemorrhage in the emergency department (SHED): a prospective, observational, multicentre cohort study

BACKGROUND

People presenting to the ED with acute severe headache often undergo investigation to exclude subarachnoid haemorrhage (SAH). International guidelines propose that brain imaging within 6 hours of headache onset can exclude SAH, in isolation. The safety of this approach is debated. We sought to externally validate this strategy and evaluate the test characteristics of CT-brain beyond 6 hours.

METHODS

A prospective, multicentre, observational cohort study of consecutive adult patients with non-traumatic acute headache presenting to the ED within a UK National Health Service setting. Investigation, diagnosis and management of SAH were all performed within routine practice. All participants were followed up for 28 days using medical records and direct contact as necessary. Uncertain diagnoses were independently adjudicated.

RESULTS

Between March 2020 and February 2023, 3663 eligible patients were enrolled from 88 EDs (mean age 45.8 (SD 16.6), 64.1% female). 3268 patients (89.2%) underwent CT-brain imaging. There were 237 cases of confirmed SAH, a prevalence of 6.5%. CT within 6 hours of headache onset (n=772) had a sensitivity of 97% (95% CI 92.5% to 99.2%) for the diagnosis of SAH and a negative predictive value of 99.6% (95% CI 98.9% to 99.9%). The post-test probability after a negative CT within 6 hours was 0.5% (95% CI 0.2% to 1.3%). The negative likelihood ratio was 0.03 (95% CI 0.01 to 0.08). CT within 24 hours of headache onset (n=2008) had a sensitivity of 94.6% (95% CI 91.0% to 97.0%). Post-test probability for SAH was consistently less than 1%. For aneurysmal SAH, post-test probability was 0.1% (95% CI 0.0% to 0.4%) if the CT was performed within 24 hours of headache onset.

CONCLUSION

Our data suggest a very low likelihood of SAH after a negative CT-brain scan performed early after headache onset. These results can inform shared decision-making on the risks and benefits of further investigation to exclude SAH in ED patients with acute headache.

Demographics and mortality trends of valvular heart disease in older adults in the United States: Insights from CDC-wonder database 1999-2019

Background

Valvular heart disease (VHD) represents a spectrum of cardiac conditions, including valvular stenosis, valvular regurgitation, or mixed lesions affecting single or multiple valves. The severity of VHD has emerged as a major cause of cardiovascular (CV) morbidity and mortality among the older population in the United States (U.S).

Objective

To evaluate temporal trends in mortality associated with VHD in the elderly U.S population between 1999 and 2019.

Methods

We utilized the CDC WONDER database for VHD mortality in adults ≥75 from 1999 to 2019, using ICD-10 codes. Age-adjusted mortality rates (AAMR) per 100,000 people with associated annual percentage change (APC) were calculated. Joinpoint regression was used to assess the overall trends and trends for demographic, geographic, and type of valvular disease subgroups.

Results

A total of 666,765 VHD deaths in older adults from 1999 to 2019 was identified, with an initial decline in AAMR until 2007 with an APC: 0.62, 95 % CI (-1.66-0.33), stability until 2014, and a significant decrease until 2019 (APC: 1.47, 95 % CI [-2.24-1.04], P < 0.0001). Men consistently had higher AAMRs compared to women (overall AAMR men: 173.6; women: 138.2). The AAMRs were found to be highest in the White (166.5), followed by American Indian or Alaska Native population at (93.8) Hispanic or Latino at (80.7), Black or African American populations at (74.1) and lastly Asian or Pacific Islander (73.4). Non-metropolitan areas manifested higher AAMRs for deaths related to VHD than metropolitan areas (overall AAMRs 160.5 vs 149.5) respectively. State-wide AAMRs varied, with the highest in Vermont at 324.2 (95 % CI [313.0-335.4], P < 0.0001) and the lowest in Mississippi at 88.0 (95 % CI [85.0-91.0], P < 0.0001). Non-rheumatic and aortic valve disorders in adults ≥75 years had higher mortality rates compared to rheumatic or mitral valve conditions in those <75 years.

Conclusion

Our study showed a decline in U.S. VHD mortality from 1999 to 2019 but found persistent disparities by gender, race, age, region, and VHD type. Targeted policies for prevention and early diagnosis are needed to address these inequalities.

Racial difference in receiving computed tomography for head injury patients in emergency departments

STUDY OBJECTIVE

Prior studies have suggested potential racial differences in receiving imaging tests in emergency departments (EDs), but the results remain inconclusive. In addition, most prior studies may only have limited racial groups for minority patients. This study aimed to investigate racial differences in head computed tomography (CT) administration rates in EDs among patients with head injuries.

METHODS

Patients with head injuries who visited EDs were examined. The primary outcome was patients receiving head CT during ED visits, and the primary exposure was patient race/ethnicity, including Asian, Hispanic, Non-Hispanic Black (Black), and Non-Hispanic White (White). Multivariable logistic regression analyses were performed using the National Hospital Ambulatory Medical Care Survey database, adjusting for patients and hospital characteristics.

RESULTS

Among 6130 patients, 51.9% received a head CT scan. Asian head injury patients were more likely to receive head CT than White patients (59.1% versus 54.0%, difference 5.1%, p < 0.001). This difference persisted in adjusted results (odds ratio, 1.52; 95% CI, 1.06-2.16, p = 0.022). In contrast, Black and Hispanic patients have no significant difference in receiving head CT than White patients after the adjustment.

CONCLUSIONS

Asian head injury patients were more likely to receive head CT than White patients. This difference may be attributed to the limited English proficiency among Asian individuals and the fact that there is a wide variety of different languages spoken by Asian patients. Future studies should examine rates of receiving other diagnostic imaging modalities among different racial groups and possible interventions to address this difference.

Acquisition parameters influence AI recognition of race in chest x-rays and mitigating these factors reduces underdiagnosis bias

A core motivation for the use of artificial intelligence (AI) in medicine is to reduce existing healthcare disparities. Yet, recent studies have demonstrated two distinct findings: (1) AI models can show performance biases in underserved populations, and (2) these same models can be directly trained to recognize patient demographics, such as predicting self-reported race from medical images alone. Here, we investigate how these findings may be related, with an end goal of reducing a previously identified underdiagnosis bias. Using two popular chest x-ray datasets, we first demonstrate that technical parameters related to image acquisition and processing influence AI models trained to predict patient race, where these results partly reflect underlying biases in the original clinical datasets. We then find that mitigating the observed differences through a demographics-independent calibration strategy reduces the previously identified bias. While many factors likely contribute to AI bias and demographics prediction, these results highlight the importance of carefully considering data acquisition and processing parameters in AI development and healthcare equity more broadly.

Audit and feedback to change diagnostic image ordering practices: A systematic review and meta-analysis

BACKGROUND

Up to 30% of diagnostic imaging (DI) tests may be unnecessary, leading to increased healthcare costs and the possibility of patient harm. The primary objective of this systematic review was to assess the effect of audit and feedback (AF) interventions directed at healthcare providers on reducing image ordering. The secondary objective was to examine the effect of AF on the appropriateness of DI ordering.

METHODS

Studies were identified using MEDLINE, EMBASE, CINAHL, Cochrane Central Register of Controlled Trials and ClinicalTrials.gov registry on December 22nd, 2022. Studies were included if they were randomized control trials (RCTs), targeted healthcare professionals, and studied AF as the sole intervention or as the core component of a multi-faceted intervention. Risk of bias for each study was evaluated using the Cochrane risk of bias tool. Meta-analyses were completed using RevMan software and results were displayed in forest plots.

RESULTS

Eleven RCTs enrolling 4311 clinicians or practices were included. AF interventions resulted in 1.5 fewer image test orders per 1000 patients seen than control interventions (95% confidence interval (CI) for the difference -2.6 to -0.4, p-value = 0.009). The effect of AF on appropriateness was not statistically significant, with a 3.2% (95% CI -1.5 to 7.7%, p-value = 0.18) greater likelihood of test orders being considered appropriate with AF vs control interventions. The strength of evidence was rated as moderate for the primary objective but was very low for the appropriateness outcome because of risk of bias, inconsistency in findings, indirectness, and imprecision.

CONCLUSION

AF interventions are associated with a modest reduction in total DI ordering with moderate certainty, suggesting some benefit of AF. Individual studies document effects of AF on image order appropriateness ranging from a non-significant trend toward worsening to a highly significant improvement, but the weighted average effect size from the meta-analysis is not statistically significant with very low certainty.

Understanding and Mitigating Bias in Imaging Artificial Intelligence

Artificial intelligence (AI) algorithms are prone to bias at multiple stages of model development, with potential for exacerbating health disparities. However, bias in imaging AI is a complex topic that encompasses multiple coexisting definitions. Bias may refer to unequal preference to a person or group owing to preexisting attitudes or beliefs, either intentional or unintentional. However, cognitive bias refers to systematic deviation from objective judgment due to reliance on heuristics, and statistical bias refers to differences between true and expected values, commonly manifesting as systematic error in model prediction (ie, a model with output unrepresentative of real-world conditions). Clinical decisions informed by biased models may lead to patient harm due to action on inaccurate AI results or exacerbate health inequities due to differing performance among patient populations. However, while inequitable bias can harm patients in this context, a mindful approach leveraging equitable bias can address underrepresentation of minority groups or rare diseases. Radiologists should also be aware of bias after AI deployment such as automation bias, or a tendency to agree with automated decisions despite contrary evidence. Understanding common sources of imaging AI bias and the consequences of using biased models can guide preventive measures to mitigate its impact. Accordingly, the authors focus on sources of bias at stages along the imaging machine learning life cycle, attempting to simplify potentially intimidating technical terminology for general radiologists using AI tools in practice or collaborating with data scientists and engineers for AI tool development. The authors review definitions of bias in AI, describe common sources of bias, and present recommendations to guide quality control measures to mitigate the impact of bias in imaging AI. Understanding the terms featured in this article will enable a proactive approach to identifying and mitigating bias in imaging AI. Published under a CC BY 4.0 license. Test Your Knowledge questions for this article are available in the supplemental material. See the invited commentary by Rouzrokh and Erickson in this issue.

Patient, Provider, and Practice Characteristics Predicting Use of Diagnostic Imaging in Primary Care: Cross-Sectional Data From the National Ambulatory Medical Care Survey

OBJECTIVE

To determine imaging utilization rates in outpatient primary care visits and factors influencing likelihood of imaging use.

METHODS

We used 2013 to 2018 National Ambulatory Medical Care Survey cross-sectional data. All visits to primary care clinics during the study period were included in the sample. Descriptive statistics on visit characteristics including imaging utilization were calculated. Logistic regression analyses evaluated the influence of a variety of patient-, provider-, and practice-level variables on the odds of obtaining diagnostic imaging, further subdivided by modality (radiographs, CT, MRI, and ultrasound). The data's survey weighting was accounted for to produce valid national-level estimates of imaging use for US office-based primary care visits.

RESULTS

Using survey weights, approximately 2.8 billion patient visits were included. Diagnostic imaging was ordered at 12.5% of visits with radiographs the most common (4.3%) and MRI the least common (0.8%). Imaging utilization was similar or greater among minority patients compared with White, non-Hispanic patients. Physician assistants used imaging at higher rates than physicians, in particular CT at 6.5% of visits compared with 0.7% for doctors of medicine and doctors of osteopathic medicine (odds ratio 5.67, 95% confidence interval 4.07-7.88).

CONCLUSION

Disparities in rates of imaging utilization for minorities seen in other health care settings were not present in this sample of primary care visits, supporting that access to primary care is a path to promote health equity. Higher rates of imaging utilization among advanced-level practitioners highlight an opportunity to evaluate imaging appropriateness and promote equitable, high-value imaging among all practitioners.

Toward fairness in artificial intelligence for medical image analysis: identification and mitigation of potential biases in the roadmap from data collection to model deployment

Purpose

To recognize and address various sources of bias essential for algorithmic fairness and trustworthiness and to contribute to a just and equitable deployment of AI in medical imaging, there is an increasing interest in developing medical imaging-based machine learning methods, also known as medical imaging artificial intelligence (AI), for the detection, diagnosis, prognosis, and risk assessment of disease with the goal of clinical implementation. These tools are intended to help improve traditional human decision-making in medical imaging. However, biases introduced in the steps toward clinical deployment may impede their intended function, potentially exacerbating inequities. Specifically, medical imaging AI can propagate or amplify biases introduced in the many steps from model inception to deployment, resulting in a systematic difference in the treatment of different groups.

Approach

Our multi-institutional team included medical physicists, medical imaging artificial intelligence/machine learning (AI/ML) researchers, experts in AI/ML bias, statisticians, physicians, and scientists from regulatory bodies. We identified sources of bias in AI/ML, mitigation strategies for these biases, and developed recommendations for best practices in medical imaging AI/ML development.

Results

Five main steps along the roadmap of medical imaging AI/ML were identified: (1) data collection, (2) data preparation and annotation, (3) model development, (4) model evaluation, and (5) model deployment. Within these steps, or bias categories, we identified 29 sources of potential bias, many of which can impact multiple steps, as well as mitigation strategies.

Conclusions

Our findings provide a valuable resource to researchers, clinicians, and the public at large.

Time ratio disparities among ED patients undergoing head CT

PURPOSE

To assess if patients who underwent head computed tomography (CT) experienced disparities in the emergency department (ED) and if the indication for head CT affected disparities.

METHODS

This study employed a retrospective, IRB-approved cohort design encompassing four hospitals. All ED patients between January 2016 and September 2020 who underwent non-contrast head CTs were included. Furthermore, key time intervals including ED length of stay (LOS), ED assessment time, image acquisition time, and image interpretation time were calculated. Time ratio (TR) was used to compare these time intervals between the groups.

RESULTS

A total of 45,177 ED visits comprising 4730 trauma cases, 5475 altered mental status cases, 11,925 cases with head pain, and 23,047 cases with other indications were included. Females had significantly longer ED LOS, ED assessment time, and image acquisition time (TR = 1.012, 1.051, 1.018, respectively, P-value < 0.05). This disparity was more pronounced in female patients with head pain complaints compared to their male counterparts (TR = 1.036, 1.059, and 1.047, respectively, P-value < 0.05). Black patients experienced significantly longer ED LOS, image acquisition time, and image assessment time (TR = 1.226, 1.349, and 1.190, respectively, P-value < 0.05). These disparities persisted regardless of head CT indications. Furthermore, patients with Medicare/Medicaid insurance also faced longer wait times in all the time intervals (TR > 1, P-value < 0.001).

CONCLUSIONS

Wait times for ED head CT completion were longer for Black patients and Medicaid/Medicare insurance holders. Additionally, females experienced extended wait times, particularly when presented with head pain complaints. Our findings underscore the importance of exploring and addressing the contributing factors to ensure equitable and timely access to imaging services in the ED.

The path toward equal performance in medical machine learning

To ensure equitable quality of care, differences in machine learning model performance between patient groups must be addressed. Here, we argue that two separate mechanisms can cause performance differences between groups. First, model performance may be worse than theoretically achievable in a given group. This can occur due to a combination of group underrepresentation, modeling choices, and the characteristics of the prediction task at hand. We examine scenarios in which underrepresentation leads to underperformance, scenarios in which it does not, and the differences between them. Second, the optimal achievable performance may also differ between groups due to differences in the intrinsic difficulty of the prediction task. We discuss several possible causes of such differences in task difficulty. In addition, challenges such as label biases and selection biases may confound both learning and performance evaluation. We highlight consequences for the path toward equal performance, and we emphasize that leveling up model performance may require gathering not only more data from underperforming groups but also better data. Throughout, we ground our discussion in real-world medical phenomena and case studies while also referencing relevant statistical theory.

Identification of White Matter Hyperintensities in Routine Emergency Department Visits Using Portable Bedside Magnetic Resonance Imaging

Background White matter hyperintensity (WMH) on magnetic resonance imaging (MRI) of the brain is associated with vascular cognitive impairment, cardiovascular disease, and stroke. We hypothesized that portable magnetic resonance imaging (pMRI) could successfully identify WMHs and facilitate doing so in an unconventional setting. Methods and Results In a retrospective cohort of patients with both a conventional 1.5 Tesla MRI and pMRI, we report Cohen's kappa (κ) to measure agreement for detection of moderate to severe WMH (Fazekas ≥2). In a subsequent prospective observational study, we enrolled adult patients with a vascular risk factor being evaluated in the emergency department for a nonstroke complaint and measured WMH using pMRI. In the retrospective cohort, we included 33 patients, identifying 16 (49.5%) with WMH on conventional MRI. Between 2 raters evaluating pMRI, the interrater agreement on WMH was strong (κ=0.81), and between 1 rater for conventional MRI and the 2 raters for pMRI, intermodality agreement was moderate (κ=0.66, 0.60). In the prospective cohort we enrolled 91 individuals (mean age, 62.6 years; 53.9% men; 73.6% with hypertension), of which 58.2% had WMHs on pMRI. Among 37 Black and Hispanic individuals, the Area Deprivation Index was higher (versus White, 51.8±12.9 versus 37.9±11.9; P<0.001). Among 81 individuals who did not have a standard-of-care MRI in the preceding year, we identified WMHs in 43 of 81 (53.1%). Conclusions Portable, low-field imaging could be useful for identifying moderate to severe WMHs. These preliminary results introduce a novel role for pMRI outside of acute care and the potential role for pMRI to reduce disparities in neuroimaging.

A Health Inequity: Associations Between Cigarette Smoking Status and Mammogram Screening Among Women of Color

INTRODUCTION

We evaluated differences in yearly mammogram screening by smoking status in a sample of US women. We also examined differences in mammogram screening by race/ethnicity, age, and health care coverage.

METHODS

Data were from 1884 women participants in the 2018 Health of Houston Survey study. Binary logistic regression was used to assess the association between smoking status (current/former/non-smokers) and mammograms within 12 months. Moderators included race/ethnicity (Hispanic, Black, Asian, Other, White), age, and health care coverage.

RESULTS

In comparison to women who were non-smokers, current and former smokers showed lower odds to get a yearly mammogram (OR = 0.720; 95% CI = 0.709, .730 and OR = 0.702; 95% CI = 0.693, 0.710, respectively). Current smokers who identified as Hispanic or Black women and former smokers who identified as Hispanic, Asian, and other women showed lower odds of getting a mammogram (OR = 0.635, 95% CI = 0.611, 0.659; OR = 0.951, 95% CI = 0.919, 0.985) and (OR = 0.663, 95% CI = 0.642, 0.684; OR = 0.282, 95% CI = 0.263, 0.302; OR = 0.548, 95% CI = 0.496, 0.606) compared to White women. There were significant interactions by age and health care coverage.

CONCLUSIONS

Women of color who are current and former smokers showed lower odds to engage in mammogram screening, thus increasing their risk of undiagnosed breast cancer when compared to non-smokers. Ethnically diverse women already experience increased health disparities and smoking puts them at exacerbated risk of health complications and death.

IMPLICATIONS

Our findings suggest that smoking status is a modifiable behavioral risk factor that requires further attention in the prevention of breast cancer in ethnic minority women. Health care institutions and policymakers need to increase their awareness of and outreach efforts to women of color who smoke. These outreach efforts should focus on increasing access to smoking interventions and cancer screenings.

Racial and Ethnic Disparities in the Imaging Workup and Treatment of Knee and Hip Osteoarthritis

BACKGROUND

There is limited evidence on sociodemographic differences in osteoarthritis management, particularly in non-African American (AA) minorities. We sought to identify differences in imaging modalities, administration of intra-articular injections, and total joint arthroplasty (TJA) between racial/ethnic groups.

METHODS

We retrospectively reviewed patients presenting to outpatient clinics with a diagnosis of hip or knee osteoarthritis from January 2013 to March 2020 at a tertiary center. Univariate analyses compared differences between groups. Multivariate logistic regression analyses determined sociodemographic predictors of imaging workup and treatment.

RESULTS

In total, 105,873 patients were included. There were 74,769 (70.6%) Caucasian, 27,117 (25.6%) AA, 1,878 (1.8%) Hispanic, 1,479 (1.4%) Asian, and 630 (0.6%) Native American patients. Multivariate analyses demonstrated that AAs had decreased odds of undergoing a knee magnetic resonance imaging (odds ratio [OR] 0.77, P < .001) or injection (OR 0.94, P = .006). Asian patients had lower odds of receiving any hip X-ray (OR 0.72, P = .047) or knee injection (OR 0.83, P = .017). AA (total knee arthroplasty [TKA]: OR 0.51, P < .001; total hip arthroplasty [THA]: OR 0.57, P < .001), Hispanic (TKA: OR 0.69, P = .003; THA: OR 0.60, P = .006), and Asian (TKA: OR 0.73, P = .010; THA: OR 0.56, P = .010) patients had lower odds of undergoing TJA compared to Caucasians. We found that higher income quartiles had greater odds of receiving a magnetic resonance imaging and TJA, males had lower odds of receiving injections and greater odds of undergoing TJA, and Medicaid and self-pay patients had lower odds of undergoing TJA (P < .05).

CONCLUSION

After adjusting for sociodemographic factors, we found disparities in the imaging, administration of injections, and/or arthroplasty for AA, Asian, and Hispanic patients. Insurance status, income, and gender were also associated with imaging and treatments performed in managing hip and knee osteoarthritis.

The Accident Environment Resulting in Fragility Fractures: A 20-year National Epidemiologic Study

INTRODUCTION

Fragility fractures are an enduring source of morbidity in the elderly with unfortunate frequency and rising costs. Although the predominant cause of fractures is generally understood to be falls, the exact stratification of the causes of fractures presenting to the emergency department has not yet been described in the literature. We sought out to stratify the primary products associated with fractures in the elderly, further describing the anatomic location of the fracture and setting of injury.

METHODS

We queried the National Electronic Injury Surveillance System database for all fractures in patients older than 65 years from January 1, 2000, to December 31, 2019. We analyzed demographic data, patient disposition, anatomic fracture location, and injury setting for the top 20 causes of fractures. Trends, proportions and distributions were analyzed using descriptive statistics.

RESULTS

A total of 901,418 visits to the Emergency Department were reviewed. Of these, 216,657 (24%) were found to have fractures. The top 20 causes for fractures accounted for a total of 173,557 (19%) fractures. The average age in our population was 80.1 years (SD 8.7). Women constituted most of the patients (127,753 [74%]). Flooring (58,347 [33.6%]) was the most common product associated with the cause of fractures, with stairs/steps (29,804 [17.2%]) and bed/bed frames (19,004 [10.9%]) being the second and third most common, respectively. Lower extremity fractures (97,195 [56%]) were more common than upper extremity fractures (63,899 [37%]). The lower trunk (pelvis, femoral neck, and lower spine) was the most common anatomic location of fractures reported (64,132 [37.0%]). Most fractures occurred either at home (113,158 [65.2%]) or at a public setting (31,162 [18.0%]).

CONCLUSIONS

Most products associated with fractures among mature adults were related to flooring, stairs, or bedding. This study offers a detailed understanding on the common products associated with fractures in mature adults and aids in discussing preventive measures for lowering fracture risk with patients, communities, and healthcare systems.

Patient Race or Ethnicity and the Use of Diagnostic Imaging: A Systematic Review

OBJECTIVE

To summarize the existing literature evaluating differences in imaging use based on patient race and ethnicity.

METHODS

The authors performed a structured search of four databases for the dates January 1, 2000, to April 13, 2021, using key words and derivatives focused on imaging and patient race. Retrieved citations were reviewed by abstract and then full text to identify articles that evaluated the likelihood of imaging use by patient race or ethnicity controlling for sociodemographic factors. Data regarding publication characteristics, study population, clinical setting, and results was extracted and summarized.

RESULTS

The structured search identified 2,938 articles of which 206 met inclusion criteria. Most studies (87%, 179 of 206) were conducted in the United States, and the majority (72%, 149 of 206) found decreased or inappropriate imaging use in minority groups. Breast cancer screening was the most common clinical setting (50%, 104 of 206), followed by cancer care (10%, 21 of 206) and general imaging use (9%, 19 of 206). Government-administered surveys were the most common data source (40%, 82 of 206). Only a small minority of studies (8%, 17 of 206) evaluated strategies to mitigate the unequal use of imaging based on patient race and ethnicity.

DISCUSSION

The existing literature shows decreased or inappropriate use of diagnostic imaging for minority patients across a wide variety of clinical settings. Although the number of articles on the topic is large, the majority are clustered around specific topics, and few articles evaluate potential strategies to reduce the inequitable use of diagnostic imaging.

Valvular Heart Disease in Relation to Race and Ethnicity: JACC Focus Seminar 4/9

Valvular heart disease (VHD) is a major global public health problem. Many regions of the world continue to grapple with the adverse consequences of untreated rheumatic heart disease, a condition that is largely preventable with timely access to diagnosis and treatment. In turn, middle- and high-income countries have experienced a rise in the prevalence of calcific aortic and mitral disease, owing in part to population aging. This public health problem is further compounded by high rates of infective endocarditis, which is associated with substantial morbidity and mortality. Yet, considerations of race and ethnicity have not taken center stage in VHD research. This is despite evidence of major health care disparities in socioeconomic and medical risk factors, access to diagnosis, and provision of appropriate treatment. In this paper, the authors review differences in the etiology, diagnosis, and treatment of VHD within the context of race, ethnicity, and health care disparities.

Revisiting racial disparities in ED CT utilization during the Affordable Care Act era: 2009-2018 data from the NHAMCS

OBJECTIVE

To examine the trends in CT utilization in the emergency department (ED) for different racial and ethnic groups, factors that may affect utilization, and the effects of increased insurance coverage since passage of the Affordable Care Act in 2010.

MATERIALS AND METHODS

Data from the National Hospital Ambulatory Medical Care Survey (NHAMCS) for the years 2009-2018 were used for the analysis. The NHAMCS is a cross-sectional survey which has random and systematical samples of more than 200,000 visits to over 250 hospital EDs in the USA. Patient demographic characteristics, source of payment/insurance, clinical presentation, and disposition from the ED were recorded. Descriptive statistics and multivariate logistic regression were performed.

RESULTS

Between 2009 and 2018, the rate of uninsured patients in the ED decreased from 18.1% to as low as 9.9%, but this was not associated with a decrease in the disparity in CT utilization between non-Hispanic Black and non-Hispanic White patients. CT use rate increased 38% over the study period. Factors strongly associated with CT utilization include age, source of payment, triage category, disposition from the ED, and residence. After controlling for these factors, non-Hispanic White patients were 21% more likely to undergo CT than non-Hispanic Black patients, though no disparity was seen for Hispanic or Asian/other groups.

CONCLUSION

Despite increased insurance coverage over the sample period, racial disparities between non-Hispanic Black and non-Hispanic White patients persist in CT utilization, though no disparity was seen for Hispanic or Asian/other patients. The source of this disparity remains unclear and is likely multifactorial.

Education of Radiologists in Healthcare Disparities

Disparities exist in access to a multitude of screening and diagnostic imaging examinations and procedures. To address these disparities within radiology, emphasis so far has been placed upon diversifying the workforce and formally educating trainees on healthcare disparities. Currently, there is no organized and nationally accepted educational program or content for practicing radiologists specific to diversity and healthcare disparity. This void can be addressed by providing an educational curriculum framework for practicing radiologists based on three key factors: individual efforts, calling for institutional change, and national collaboration. Individual efforts should focus on acknowledging the existence of disparities, understanding the contribution of one's implicit bias in perpetuating disparities, understanding and highlighting issues related to insurance coverage of radiology examinations, and participating in radiology political action committees. These efforts can be facilitated by a consolidated web-based training program for practicing radiologists. To pave the way for meaningful systemic change, the implementation of institutional change like that initiated by the Culture of Safety movement in 2002 is needed. A national collaborative effort initiated by radiology organizations to empower radiologists and recognize positive changes would further provide support. SUMMARY: A three-pronged educational framework combining individual radiologist education, institutional change, and national collaboration will enable radiologists to play a role in addressing imaging-related disparities in healthcare.

Reducing Health Disparities in Radiology Through Social Determinants of Health: Lessons From the COVID-19 Pandemic

During the COVID-19 pandemic, the disproportionate morbidity and mortality borne by racial minorities, patients of lower socioeconomic status, and patients lacking health insurance reflect the critical role of social determinants of health, which are manifestations of entrenched structural inequities. In radiology, social determinants of health lead to disparate use of imaging services through multiple intersecting contributors, on both the provider and patient side, affecting diagnosis and treatment. Disparities on the provider side include ordering of initial or follow-up imaging studies and providing standard-of-care interventional procedures, while patient factors include differences in awareness of screening exams and confidence in the healthcare system. Disparate utilization of mammography and lung cancer screening lead to delayed diagnosis, while differential provision of minimally invasive interventional procedures contributes to differential outcomes related to treatment. Interventions designed to mitigate social determinants of health could help to equalize the healthcare system. Here we review disparities in access and health outcomes in radiology. We investigate underlying contributing factors in order to identify potential policy changes that could promote more equitable health in radiology.

Narrowing the Gap: Imaging Disparities in Radiology

It may seem unlikely that the field of radiology perpetuates disparities in health care, as most radiologists never interact directly with patients, and racial bias is not an obvious factor when interpreting images. However, a closer look reveals that imaging plays an important role in the propagation of disparities. For example, many advanced and resource-intensive imaging modalities, such as MRI and PET/CT, are generally less available in the hospitals frequented by people of color, and when they are available, access is impeded due to longer travel and wait times. Furthermore, their images may be of lower quality, and their interpretations may be more error prone. The aggregate effect of these imaging acquisition and interpretation disparities in conjunction with social factors is insufficiently recognized as part of the wide variation in disease outcomes seen between races in America. Understanding the nature of disparities in radiology is important to effectively deploy the resources and expertise necessary to mitigate disparities through diversity and inclusion efforts, research, and advocacy. In this article, the authors discuss disparities in access to imaging, examine their causes, and propose solutions aimed at addressing these disparities.

Patient Race/Ethnicity and Diagnostic Imaging Utilization in the Emergency Department: A Systematic Review

PURPOSE

Diagnostic imaging often is a critical contributor to clinical decision making in the emergency department (ED). Racial and ethnic disparities are widely reported in many aspects of health care, and several recent studies have reported a link between patient race/ethnicity and receipt of imaging in the ED.

METHODS

The authors conducted a systematic review following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, searching three databases (PubMed, Embase, and the Cochrane Library) through July 2020 using keywords related to diagnostic imaging, race/ethnicity, and the ED setting, including both adult and pediatric populations and excluding studies that did not control for the important confounders of disease severity and insurance status.

RESULTS

The search strategy identified 7,313 articles, of which 5,668 underwent title and abstract screening and 238 full-text review, leaving 42 articles meeting the inclusion criteria. Studies were predominately conducted in the United States (41), split between adult (13) and pediatric (17) populations or both (12), and spread across a variety of topics, mostly focusing on specific anatomic regions or disease processes. Most studies (30 of 42 [71.4%]) reported an association between Black, African American, Hispanic, or nonwhite race/ethnicity and decreased receipt of imaging.

CONCLUSIONS

Despite heterogeneity among studies, patient race/ethnicity is linked with receipt of diagnostic imaging in the ED. The strength and directionality of this association may differ by specific subpopulation and disease process, and more efforts to understand potential underlying factors are needed.