Article

Clinical Implications of Removing Race From Estimates of Kidney Function

December 2020
JAMA The Journal of the American Medical Association 325(2)

DOI:10.1001/jama.2020.22124

Authors:

Herman Taylor

Morehouse School of Medicine

J. Kevin Tucker

Brigham and Women's Hospital

Show all 7 authorsHide

Over the past year, medical centers across the US have removed race adjustment from estimated glomerular filtration rate from serum creatinine (eGFRcr), with many now reporting the “White/other” value for all patients. These changes follow calls to reconsider the use of race in estimating kidney function¹ and in medicine broadly.² We analyzed potential changes in recommended care using eGFRcr with and without race among Black individuals in the US (individuals who are not Black would not be affected).

Development and validation of a race-agnostic computable phenotype for kidney health in adult hospitalized patients

Article

Full-text available

Apr 2024
PLOS ONE

Standard race adjustments for estimating glomerular filtration rate (GFR) and reference creatinine can yield a lower acute kidney injury (AKI) and chronic kidney disease (CKD) prevalence among African American patients than non–race adjusted estimates. We developed two race-agnostic computable phenotypes that assess kidney health among 139,152 subjects admitted to the University of Florida Health between 1/2012–8/2019 by removing the race modifier from the estimated GFR and estimated creatinine formula used by the race-adjusted algorithm (race-agnostic algorithm 1) and by utilizing 2021 CKD-EPI refit without race formula (race-agnostic algorithm 2) for calculations of the estimated GFR and estimated creatinine. We compared results using these algorithms to the race-adjusted algorithm in African American patients. Using clinical adjudication, we validated race-agnostic computable phenotypes developed for preadmission CKD and AKI presence on 300 cases. Race adjustment reclassified 2,113 (8%) to no CKD and 7,901 (29%) to a less severe CKD stage compared to race-agnostic algorithm 1 and reclassified 1,208 (5%) to no CKD and 4,606 (18%) to a less severe CKD stage compared to race-agnostic algorithm 2. Of 12,451 AKI encounters based on race-agnostic algorithm 1, race adjustment reclassified 591 to No AKI and 305 to a less severe AKI stage. Of 12,251 AKI encounters based on race-agnostic algorithm 2, race adjustment reclassified 382 to No AKI and 196 (1.6%) to a less severe AKI stage. The phenotyping algorithm based on refit without race formula performed well in identifying patients with CKD and AKI with a sensitivity of 100% (95% confidence interval [CI] 97%–100%) and 99% (95% CI 97%–100%) and a specificity of 88% (95% CI 82%–93%) and 98% (95% CI 93%–100%), respectively. Race-agnostic algorithms identified substantial proportions of additional patients with CKD and AKI compared to race-adjusted algorithm in African American patients. The phenotyping algorithm is promising in identifying patients with kidney disease and improving clinical decision-making.

Evaluating and Improving the Performance and Racial Fairness of Algorithms for GFR Estimation

Preprint

Full-text available

Jan 2024

Data-driven clinical prediction algorithms are used widely by clinicians. Understanding what factors can impact the performance and fairness of data-driven algorithms is an important step towards achieving equitable healthcare. To investigate the impact of modeling choices on the algorithmic performance and fairness, we make use of a case study to build a prediction algorithm for estimating glomerular filtration rate (GFR) based on the patient’s electronic health record (EHR). We compare three distinct approaches for estimating GFR: CKD-EPI equations, epidemiological models, and EHR-based models. For epidemiological models and EHR-based models, four machine learning models of varying computational complexity (i.e., linear regression, support vector machine, random forest regression, and neural network) were compared. Performance metrics included root mean squared error (RMSE), median difference, and the proportion of GFR estimates within 30% of the measured GFR value (P30). Differential performance between non-African American and African American group was used to assess algorithmic fairness with respect to race. Our study showed that the variable race had a negligible effect on error, accuracy, and differential performance. Furthermore, including more relevant clinical features (e.g., common comorbidities of chronic kidney disease) and using more complex machine learning models, namely random forest regression, significantly lowered the estimation error of GFR. However, the difference in performance between African American and non-African American patients did not decrease, where the estimation error for African American patients remained consistently higher than non-African American patients, indicating that more objective patient characteristics should be discovered and included to improve algorithm performance.

Race-Modified Equations Estimating Renal Function and the Disparity in Partial Nephrectomy Use in Black Patients

Article

Mar 2023

Purpose Black patients with renal tumors are less likely to undergo partial nephrectomy (PN) despite a higher risk of chronic kidney disease (CKD). Racialized equations estimating glomerular filtration rate (eGFR) estimate a 15% to 20% higher renal function for Black patients as opposed to non-Black patients. Whether raced-modified eGFR equations affect the use of nephron-sparing surgery in Black patients has not yet been explored. Materials and Methods This is a single-center, retrospective review of 6327 consecutive patients who underwent PN (n = 3533) or radical nephrectomy (RN) (n = 2794) from 2005 to 2020. Patients were asked to self-report their race preoperatively. We excluded patients with a tumor thrombus (n = 149). Data were dichotomized into Black patients (n = 654) and non-Black patients (n = 5673). Multivariable logistic regression analysis controlled for known factors associated with PN. The impact of the race-modified (2009CKD-EPI) and race-free (2021CKD-EPI-refit) equations on the distribution of patients across CKD stages was analyzed. Results Among 6327 patients, 10.3% were Black patients. Black patients had higher comorbidity rates and lower preoperative eGFR than non-Black patients ( P < .001). Black patients had lower overall PN rates than non-Black patients (49% vs 57%; P < .001), a finding that was most pronounced in the CKD stage 1 group (55.84% vs 67.43%; P = .0011). On multivariable logistic regression analysis, the Black race was associated with lower odds of PN (OR = 0.76 (0.61-0.96), P = .001). CKD-EPI-refit moved 27% of Black patients to a lower and more accurate CKD stage. Conclusions Using a conventional race-modified equation potentially overestimated the renal function of 27% of Black patients who were restaged into worse CKD stages when using the race-free equation. While the surgical approach is multifactorial, race-modified equations may overestimate renal function in a significant portion of Black patients and explain some, but not all, of the observed lower utilization of nephron-sparing surgical approaches in Black patients.

The American Society for Bone and Mineral Research Task Force on clinical algorithms for fracture risk report

Article

Apr 2024

Using race and ethnicity in clinical algorithms potentially contributes to health inequities. The American Society for Bone and Mineral Research (ASBMR) Professional Practice Committee convened the ASBMR Task Force on Clinical Algorithms for Fracture Risk to determine the impact of race and ethnicity adjustment in the US Fracture Risk Assessment Tool (US-FRAX). The Task Force engaged the University of Minnesota Evidence–based Practice Core to conduct a systematic review investigating the performance of US-FRAX for predicting incident fractures over 10 years in Asian, Black, Hispanic, and White individuals. Six studies from the Women’s Health Initiative (WHI) and Study of Osteoporotic Fractures (SOF) were eligible; cohorts only included women and were predominantly White (WHI > 80% and SOF > 99%), data were not consistently stratified by race and ethnicity, and when stratified there were far fewer fractures in Black and Hispanic women vs White women rendering area under the curve (AUC) estimates less stable. In the younger WHI cohort (n = 64 739), US-FRAX without bone mineral density (BMD) had limited discrimination for major osteoporotic fracture (MOF) (AUC 0.53 (Black), 0.57 (Hispanic), and 0.57 (White)); somewhat better discrimination for hip fracture in White women only (AUC 0.54 (Black), 0.53 (Hispanic), and 0.66 (White)). In a subset of the older WHI cohort (n = 23 918), US-FRAX without BMD overestimated MOF. The Task Force concluded that there is little justification for estimating fracture risk while incorporating race and ethnicity adjustments and recommends that fracture prediction models not include race or ethnicity adjustment but instead be population-based and reflective of US demographics, and inclusive of key clinical, behavioral, and social determinants (where applicable). Research cohorts should be representative vis-à-vis race, ethnicity, gender, and age. There should be standardized collection of race and ethnicity; collection of social determinants of health to investigate impact on fracture risk; and measurement of fracture rates and BMD in cohorts inclusive of those historically underrepresented in osteoporosis research.

Race‐free renal function estimation equations and potential impact on Black patients: Implications for cancer clinical trial enrollment

Article

Full-text available

Jan 2023
CANCER-AM CANCER SOC

Background Black patients face disparities in cancer outcomes. Additionally, Black patients are more likely to be undertreated and underrepresented in clinical trials. The recent recommendation to remove race from the estimated glomerular filtration rate (eGFR) results in lower eGFR values for Black patients. The ramifications of this decision, both intended and unintended, are still being elucidated in the medical community. Here, the authors analyze the removal of race from eGFR for Black patients with cancer, specifically with respect to clinical trial eligibility. Methods In a cohort of self‐identified Black patients who underwent nephrectomy at a tertiary referral center from 2009 to 2021 (n = 459), eGFR was calculated with and without race in commonly used equations (Chronic Kidney Disease Epidemiology Collaboration [CKD‐EPI] and Modification of Diet in Renal Disease [MDRD]). The distribution of patients and changes within chronic kidney disease stages with different equations was considered. Theoretical exclusion at commonly observed clinical trial eGFR points was then simulated on the basis of the utilization of the race coefficient. Results The median eGFR from CKD‐EPI was significantly higher with race (76 ml/min/1.73 m²) than without race (66 ml/min/1.73 m²; p < .0001). The median eGFR from MDRD was significantly higher with race (71.0 ml/min/1.73 m²) than without race (58 ml/min/1.73 m²; p < .0001). Observing results in the context of common clinical trial cutoff points, the authors found that 13%–22%, 6%–12%, and 2%–3% more Black patients would fall under common clinical trial cutoffs of 60, 45, and 30 ml/min, respectively, depending on the equation used. A subanalysis of stage III–IV patients only was similar. Conclusions Race‐free renal function equations may inadvertently result in increased exclusion of Black patients from clinical trials. This is especially concerning because of the underrepresentation and undertreatment that Black patients already experience. Plain Language Summary Black patients experience worse oncologic outcomes and are underrepresented in clinical trials. Kidney function, as estimated by glomerular filtration rate equations, is a factor in who can and cannot be in a clinical trial. Race is a variable in some of these equations. For Black patients, removing race from these equations leads to the calculation of lower kidney function. Lower estimated kidney function may result in more black patients being excluded from clinical trials. The inclusion of all races in clinical trials is important for offering best care to everyone and for making results from clinical trials applicable to everyone.

New analysis of the comparative diagnostic performances of five different eGFR equations: The revised 2021 CKD-EPI eGFR, and the four older equations

Article

Full-text available

Dec 2022

Purpose: In 2021, the CKD-EPI-creatinine and CKD-EPI-creatinine-cystatin-C combined equations were revised again by the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) and the National Institute of Diabetes & Digestive and Kidney Diseases (NIDDK) and it was reported that its accuracy was increased. The main object of this study is to compare the diagnostic accuracy and performance of 2009 CKD-EPI eGFRcr, 2012 CKD-EPI eGFRCysC, 2012 CKD-EPI eGFRcr-CysC, MDRD, and revised 2021 CKD-EPI to provide a better estimation. Materials and Methods: The study design was prospective. The sample consisted of 111 CKD patients and 35 healthy individuals who applied to the nephrology clinic. All participants were evaluated by a nephrologist. The participants were divided into study groups according to their KDIGO classification. Five variables were used: age, race, gender, serum creatinine, and serum cystatin-C. Results: 2012 CKD-EPI eGFR cystatin-C out performed existing equations in terms of accuracy, specificity, and sensitivity (AUC:0.988). Furthermore, when both creatinine and cystatin-C are included in an equation, it consistently improves diagnostic values over formulas containing only creatinine (2009CKD-EPI eGFRcr AUC:0.953, 2012CKD-EPI eGFRcr-CysC AUC: 0.985, 2021CKD-EPI eGFRcr AUC:0.954, MDRD AUC: 0.953, 2021CKD-EPI eGFRcr-CysC AUC: 0.985). Conclusion: The new formula has not been confirmed to be superior to other equations in its ability to estimate eGFR values, particularly at higher levels of chronic kidney disease. When compared with existing equations, 2012 CKD-EPI eGFRCysC had higher specificity and sensitivity. Combined formulas containing both blood tests had higher diagnostic values than equations containing only creatinine. Our results can help inform which equations to use for better estimation in renal function screening.

National Projections for Clinical Implications of Race-Free Creatinine-Based GFR Estimating Equations

Article

Nov 2022

Background The National Kidney Foundation and American Society of Nephrology Task Force on Reassessing the Inclusion of Race in Diagnosing Kidney Disease recently recommended a new race-free creatinine-based equation for eGFR. The effect on recommended clinical care across race and ethnicity groups is unknown. Methods We analyzed nationally representative cross-sectional questionnaires and medical examinations from 44,360 participants collected between 2001 and 2018 by the National Health and Nutrition Examination Survey. We quantified the number and proportion of Black, White, Hispanic, and Asian/Other adults with guideline-recommended changes in care. Results The new equation, if applied nationally, could assign new CKD diagnoses to 434,000 (95% confidence interval [CI], 350,000 to 517,000) Black adults, reclassify 584,000 (95% CI, 508,000 to 667,000) to more advanced stages of CKD, restrict kidney donation eligibility for 246,000 (95% CI, 189,000 to 303,000), expand nephrologist referrals for 41,800 (95% CI, 19,800 to 63,800), and reduce medication dosing for 222,000 (95% CI, 169,000 to 275,000). Among non-Black adults, these changes may undo CKD diagnoses for 5.51 million (95% CI, 4.86 million to 6.16 million), reclassify 4.59 million (95% CI, 4.28 million to 4.92 million) to less advanced stages of CKD, expand kidney donation eligibility for 3.96 million (95% CI, 3.46 million to 4.46 million), reverse nephrologist referral for 75,800 (95% CI, 35,400 to 116,000), and reverse medication dose reductions for 1.47 million (95% CI, 1.22 million to 1.73 million). The racial and ethnic mix of the populations used to develop eGFR equations has a substantial effect on potential care changes. Conclusion The newly recommended 2021 CKD-EPI creatinine-based eGFR equation may result in substantial changes to recommended care for US patients of all racial and ethnic groups.

Impact of race-based calculations of eGFR on the management of muscle invasive bladder cancer

Article

Dec 2024

Amir Khan

A Framework for Considering the Value of Race and Ethnicity in Estimating Disease Risk

Article

Dec 2024
ANN INTERN MED

Background: Accounting for race and ethnicity in estimating disease risk may improve the accuracy of predictions but may also encourage a racialized view of medicine. Objective: To present a decision analytic framework for considering the potential benefits of race-aware over race-unaware risk predictions, using cardiovascular disease, breast cancer, and lung cancer as case studies. Design: Cross-sectional study. Setting: NHANES (National Health and Nutrition Examination Survey), 2011 to 2018, and NLST (National Lung Screening Trial), 2002 to 2004. Patients: U.S. adults. Measurements: Starting with risk predictions from clinically recommended race-aware models, the researchers generated race-unaware predictions via statistical marginalization. They then estimated the utility gains of the race-aware over the race-unaware models, based on a simple utility function that assumes constant costs of screening and constant benefits of disease detection. Results: The race-unaware predictions were substantially miscalibrated across racial and ethnic groups compared with the race-aware predictions as the benchmark. However, the clinical net benefit at the population level of race-aware predictions over race-unaware predictions was smaller than expected. This result stems from 2 empirical patterns: First, across all 3 diseases, 95% or more of individuals would receive the same decision regardless of whether race and ethnicity are included in risk models; second, for those who receive different decisions, the net benefit of screening or treatment is relatively small because these patients have disease risks close to the decision threshold (that is, the theoretical “point of indifference”). When used to inform rationing, race-aware models may have a more substantial net benefit. Limitations: For illustrative purposes, the race-aware models were assumed to yield accurate estimates of risk given the input variables. The researchers used a simplified approach to generate race-unaware risk predictions from the race-aware models and a simple utility function to compare models. Conclusion: The analysis highlights the importance of foregrounding changes in decisions and utility when evaluating the potential benefit of using race and ethnicity to estimate disease risk.

Strategies to advancing digital health equity: perspectives from underserved groups

Conference Paper

Oct 2024

Digital Health Technologies (DHTs) are revolutionizing healthcare. However, there is a lack of demonstrable health benefits across all populations. To advance digital health equity, we explored the perspectives of underserved groups on strategies to support digital inclusivity. Participants belonged to two or more CLEARS (Culture (ethnicity, language, religion), Limiting conditions (visual or hearing), low Educational attainment, older Age, Residence (rural, deprived or no fixed address), low Socioeconomic status) groups. This framework encompassed sociodemographic factors associated with digital exclusion and recognizes the role of intersectionality.(1) Twenty-nine individuals participated in focus groups (n = 4) or interviews (n = 11) and shared their perspectives on the barriers to DHTs, innovative ideas to improve inclusivity, and perceived feasibility of implementing changes. An inductive reflective thematic analysis was used to analyze the transcripts, assisted by N-Vivo. Participants experienced fragmented communication with healthcare providers when using DHTs, explaining how they received a no-reply SMS message asking them to confirm an appointment via a link but were unable to access the link. Current solutions to overcoming such barriers included using non-digital routes (e.g., visiting a pharmacy) or calling non-emergency healthcare services. Many participants expressed the need for DHT support services to be implemented to provide educational support and access to devices. However, some participants raised concerns about the financial strain this might place on healthcare systems. Future research should evaluate the feasibility of implementing different DHTs support services across different settings to advance digital health equity. (1)Wilson et al. (2023). Strategies to develop, design, implement and support digital health equity in key underserved groups: protocol for a systematic review. Alzheimer’s & Dementia. Key messages • Strategies to mitigating digital exclusion within healthcare should be designed with underserved groups. • Future research should evaluate the feasibility of implementing different DHTs support services across different settings to advance digital health equity.

A systematic review of fairness in machine learning

Article

Full-text available

Sep 2024

Fairness in Machine Learning (ML) has emerged as a crucial concern as these models increasingly influence critical decisions in various domains, including healthcare, finance, and criminal justice. The presence of bias in ML systems can lead to unfair and discriminatory outcomes, undermining the reliability and ethical standards of these technologies. As the deployment of ML expands, ensuring that these systems are fair and unbiased is not only a technical challenge but also a moral imperative. Here, a systematic literature review was conducted to explore fairness in machine learning, utilizing the ACM, IEEE, and Springer databases. From an initial retrieval of 975 papers, 30 were included in the review. The results highlight the identification of sensitive attributes, the metrics used to assess bias, and the various databases tested. Additionally, the review categorizes the in-processing and post-processing approaches employed to mitigate bias and examines how studies are managing the trade-off between fairness and accuracy. This comprehensive analysis provides a detailed understanding of the current state of fairness in machine learning and offers insights into effective strategies for bias mitigation.

Race as a Risk Marker, Not a Risk Factor: Revising Race-Based Algorithms to Protect Racially Oppressed Patients

Article

Full-text available

Jul 2024
J GEN INTERN MED

Emerging consensus in the medical and public health spheres encourages removing race and ethnicity from algorithms used in clinical decision-making. Although clinical algorithms remain appealing given their promise to lighten the cognitive load of medical practice and save time for providers, they risk exacerbating existing health disparities. Race is a strong risk marker of health outcomes, yet it is not a risk factor. The use of race as a factor in medical algorithms suggests that the effect of race is intrinsic to the patient or that its effects can be distinct or separated from other social and environmental variables. By contrast, incisive public health analysis coupled with a race-conscious perspective recognizes that race serves as a marker of countless other dynamic variables and that structural racism, rather than race, compromises the health of racially oppressed individuals. This perspective offers a historical and theoretical context for the current debates regarding the use of race in clinical algorithms, clinical and epidemiologic perspectives on “risk,” and future directions for research and policy interventions that combat color-evasive racism and follow the principles of race-conscious medicine.

Race-modified estimated glomerular filtration rate underestimates chronic kidney disease prevalence in Black patients undergoing partial and radical nephrectomy: Implications for surgical planning

Article

Jul 2024

Introduction In estimated glomerular filtration rate equations (eGFR), the race multiplier (RM) yields greater eGFR values and may assign less severe chronic kidney disease (CKD) stages to black individuals. When deciding on appropriateness for partial nephrectomy (PN), patients with CKD are often considered a relative or absolute indication. We hypothesize that the eGFR RM may have ramifications for patients being counseled for radical nephrectomy (RN) versus PN to manage their renal tumor. Methods We utilized prospective and retrospective, IRB-approved single-center databases to select patients who underwent PN or RN between 2016 and 2022. Demographics, preoperative risk factors, preoperative eGFR, and surgical management were collected. Descriptive statistics and two-tailed difference of proportion tests compared the percentage of patients with CKD who underwent nephrectomy. Results This cohort included 1137 patients who underwent RN or PN, including 74 (6.5%) Black patients and 93.5% ( n = 1063) non-Black patients. There was no statistically significant difference between the eGFR of Black and non-Black individuals using the Modification of Diet in Renal Disease equation ( P = 0.24) or Chronic Kidney Disease Epidemiology Collaboration 2009 (CKD-EPI 2009) ( P = 0.45); however, there was statistically significant difference in eGFR between sample populations when using CKD-EPI 2021 ( P = 0.0055). Of the Black patient cohort, 16.2% of patients reclassified to a worse CKD class using CKD-EPI 2021, including 9.5% of Black patients reclassified to CKD3a or worse, and 14.6% of all patients (Black and non-Black) reclassified to a different CKD class under the CKD-EPI 2021 equation. Conclusions There are quantitative differences in the evaluation of eGFR when utilizing different equations that may impact clinical considerations and health equity outcomes for nephrectomy across racial groups.

Recommendations to advance digital health equity: a systematic review of qualitative studies

Article

Full-text available

Jun 2024

The World Health Organisation advocates Digital Health Technologies (DHTs) for advancing population health, yet concerns about inequitable outcomes persist. Differences in access and use of DHTs across different demographic groups can contribute to inequities. Academics and policy makers have acknowledged this issue and called for inclusive digital health strategies. This systematic review synthesizes literature on these strategies and assesses facilitators and barriers to their implementation. We searched four large databases for qualitative studies using terms relevant to digital technology, health inequities, and socio-demographic factors associated with digital exclusion summarised by the CLEARS framework (Culture, Limiting conditions, Education, Age, Residence, Socioeconomic status). Following the PRISMA guidelines, 10,401 articles were screened independently by two reviewers, with ten articles meeting our inclusion criteria. Strategies were grouped into either outreach programmes or co-design approaches. Narrative synthesis of these strategies highlighted three key themes: firstly, using user-friendly designs, which included software and website interfaces that were easy to navigate and compatible with existing devices, culturally appropriate content, and engaging features. Secondly, providing supportive infrastructure to users, which included devices, free connectivity, and non-digital options to help access healthcare. Thirdly, providing educational support from family, friends, or professionals to help individuals develop their digital literacy skills to support the use of DHTs. Recommendations for advancing digital health equity include adopting a collaborative working approach to meet users’ needs, and using effective advertising to raise awareness of the available support. Further research is needed to assess the feasibility and impact of these recommendations in practice.

Urological impact of race-free estimated glomerular filtration rate equations

Article

Full-text available

Dec 2023

Race, Healthcare, and Health Disparities: A Critical Review and Recommendations for Advancing Health Equity

Article

Full-text available

Sep 2023

An overwhelming body of evidence points to an inextricable link between race and health disparities in the United States. Although race is best understood as a social construct, its role in health outcomes has historically been attributed to increasingly debunked theories of underlying biological and genetic differences across races. Recently, growing calls for health equity and social justice have raised awareness of the impact of implicit bias and structural racism on social determinants of health, healthcare quality, and ultimately, health outcomes. This more nuanced recognition of the role of race in health disparities has, in turn, facilitated introspective racial disparities research, root cause analyses, and changes in practice within the medical community. Examining the complex interplay between race, social determinants of health, and health outcomes allows systems of health to create mechanisms for checks and balances that mitigate unfair and avoidable health inequalities. As one of the specialties most intertwined with social medicine, emergency medicine (EM) is ideally positioned to address racism in medicine, develop health equity metrics, monitor disparities in clinical performance data, identify research gaps, implement processes and policies to eliminate racial health inequities, and promote anti-racist ideals as advocates for structural change. In this critical review our aim was to (a) provide a synopsis of racial disparities across a broad scope of clinical pathology interests addressed in emergency departments—communicable diseases, non-communicable conditions, and injuries—and (b) through a race-conscious analysis, develop EM practice recommendations for advancing a culture of equity with the potential for measurable impact on healthcare quality and health outcomes.

Heart-Kidney Allocation Policy and Proposals Impact Equity of Access for Patients with End-Stage Kidney Disease

Article

Full-text available

Aug 2023

Purpose of Review To summarize contemporary evidence describing the most appropriate candidates for simultaneous heart kidney transplantation (SHK), estimates of how SHK affects the entire transplantation system and implications of recently enacted SHK allocation policy on kidney transplantation. Recent Findings SHK has benefits for both dialysis-dependent patients and non-dialysis-dependent patients with eGFRs up to approximately 40 mL/min/1.73 m². Additionally, there is emerging data that patients with left ventricular assist devices may be inferior candidates for SHK. Recent modelling has also shown that multiorgan transplant generally diverts kidneys away from patients waiting for kidney transplant alone, who suffer inferior wait list outcomes relative to individuals who receive kidney offers from a heart, liver, or lung donor. Summary SHK is an important part of the transplant system, but increased scrutiny is warranted. Further investigations into optimal candidate selection, preoperative and perioperative reno-protective strategies, and the effects of new SHK policies on the kidney alone waitlisted population are needed.

Digital Endpoints for Assessing Instrumental Activities of Daily Living in Mild Cognitive Impairment: Systematic Review (Preprint)

Preprint

Jan 2023

BACKGROUND Subtle impairments in instrumental activities of daily living (IADLs) can be a key predictor of disease progression and are considered central to functional independence. Mild cognitive impairment (MCI) is a syndrome associated with significant changes in cognitive function and mild impairment in complex functional abilities. The early detection of functional decline through the identification of IADL impairments can aid early intervention strategies. Digital health technology is an objective method of capturing IADL-related behaviors. However, it is unclear how these IADL-related behaviors have been digitally assessed in the literature and what differences can be observed between MCI and normal aging. OBJECTIVE This review aimed to identify the digital methods and metrics used to assess IADL-related behaviors in people with MCI and report any statistically significant differences in digital endpoints between MCI and normal aging and how these digital endpoints change over time. METHODS A total of 16,099 articles were identified from 8 databases (CINAHL, Embase, MEDLINE, ProQuest, PsycINFO, PubMed, Web of Science, and Scopus), out of which 15 were included in this review. The included studies must have used continuous remote digital measures to assess IADL-related behaviors in adults characterized as having MCI by clinical diagnosis or assessment. This review was conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. RESULTS Ambient technology was the most commonly used digital method to assess IADL-related behaviors in the included studies (14/15, 93%), with passive infrared motion sensors (5/15, 33%) and contact sensors (5/15, 33%) being the most prevalent types of methods. Digital technologies were used to assess IADL-related behaviors across 5 domains: activities outside of the home, everyday technology use, household and personal management, medication management, and orientation . Other recognized domains— culturally specific tasks and socialization and communication —were not assessed. Of the 79 metrics recorded among 11 types of technologies, 65 (82%) were used only once. There were inconsistent findings around differences in digital IADL endpoints across the cognitive spectrum, with limited longitudinal assessment of how they changed over time. CONCLUSIONS Despite the broad range of metrics and methods used to digitally assess IADL-related behaviors in people with MCI, several IADLs relevant to functional decline were not studied. Measuring multiple IADL-related digital endpoints could offer more value than the measurement of discrete IADL outcomes alone to observe functional decline. Key recommendations include the development of suitable core metrics relevant to IADL-related behaviors that are based on clinically meaningful outcomes to aid the standardization and further validation of digital technologies against existing IADL measures. Increased longitudinal monitoring is necessary to capture changes in digital IADL endpoints over time in people with MCI. CLINICALTRIAL PROSPERO International Prospective Register of Systematic Reviews CRD42022326861; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=326861

Key considerations when developing and implementing digital technology for early detection of dementia-causing diseases amongst healthcare professionals: a qualitative study (Preprint)

Article

Full-text available

Feb 2023
J MED INTERNET RES

Background The World Health Organization (WHO) promotes using digital technologies to accelerate global attainment of health and well-being. This has led to a growth in research exploring the use of digital technology to aid early detection and preventative interventions for dementia-causing diseases such as Alzheimer disease. The opinions and perspectives of health care professionals must be incorporated into the development and implementation of technology to promote its successful adoption in clinical practice. Objective This study aimed to explore health care professionals’ perspectives on the key considerations of developing and implementing digital technologies for the early detection of dementia-causing diseases in the National Health Service (NHS). Methods Health care professionals with patient-facing roles in primary or secondary care settings in the NHS were recruited through various web-based NHS clinical networks. Participants were interviewed to explore their experiences of the current dementia diagnostic practices, views on early detection and use of digital technology to aid these practices, and the challenges of implementing such interventions in health care. An inductive thematic analysis approach was applied to identify central concepts and themes in the interviews, allowing the data to determine our themes. A list of central concepts and themes was applied systematically to the whole data set using NVivo (version 1.6.1; QSR International). Using the constant comparison technique, the researchers moved backward and forward between these data and evolving explanations until a fit was made. ResultsEighteen semistructured interviews were conducted, with 11 primary and 7 secondary care health care professionals. We identified 3 main categories of considerations relevant to health care service users, health care professionals, and the digital health technology itself. Health care professionals recognized the potential of using digital technology to collect real-time data and the possible benefits of detecting dementia-causing diseases earlier if an effective intervention were available. However, some were concerned about postdetection management, questioning the point of an early detection of dementia-causing diseases if an effective intervention cannot be provided and feared this would only lead to increased anxiety in patients. Health care professionals also expressed mixed opinions on who should be screened for early detection. Some suggested it should be available to everyone to mitigate the chance of excluding those who are not in touch with their health care or are digitally excluded. Others were concerned about the resources that would be required to make the technology available to everyone. Conclusions This study highlights the need to design digital health technology in a way that is accessible to all and does not add burden to health care professionals. Further work is needed to ensure inclusive strategies are used in digital research to promote health equity.

Racial and Ethnic Demographic Reporting in Phase 2 Proton Therapy Clinical Trials: A Review

Article

Full-text available

May 2023

Purpose Equitable inclusion of racial and ethnic participation in clinical trials is crucial to improving disparities in health care, especially for historically marginalized populations. Our study aims to describe the racial and ethnic demographics of patients enrolled in published phase 2 clinical trials involving proton therapy in the United States. Materials and Methods Published manuscripts were identified in PubMed, Embase, World of Science, and Cochrane. Phase 2 trials evaluating proton therapy for US patients were included. For each article in the study, data were collected comprising authors, title, and publication year, and clinical trial numbers were verified. Additional data included tumor site, primary institution, sample size, reported race/ethnicity, and raw number/percentile of race/ethnicity. Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines were used. Results Overall, 970 titles were identified; 636 remained after duplicate screening, and 75 full-text articles were assessed. We identified 38 eligible manuscripts for inclusion comprising 2648 patients. Only 15 (39%) of the publications reported race/ethnicity. Of these, 8 (21%) and 10 (26%) documented Hispanic or Black trial participants, respectively; however, only 6 (16%) documented trial participation for both Hispanic and Black patients. Of the 1409 patients with a documented race/ethnicity, 89.0% (n = 1254) were non-Hispanic white, 5.3% (n = 75) were Black, and 2.2% (n = 31) were Hispanic. Other and unknown race/ethnicity comprised the remaining patients (3.5%; n = 49). Conclusion We identified underreporting of demographic data in published phase 2 proton therapy trials, which unfortunately mirrored underreporting for cancer drug clinical trials. We also noted dramatic Black and Hispanic patient underrepresentation across the trials in which race and ethnicity are reported. Findings highlight the urgent need to identify and address barriers to proton therapy trials for Black and Hispanic patients ensuring clinical trials in radiation oncology are representative of the patients seen in clinical practice.

The impact of a student-led anti-racism programme on medical students’ perceptions and awareness of racial bias in medicine and confidence to advocate against racism

Article

Full-text available

Feb 2023
Med Educ Online

Introduction Systemic racism impacts personal and community health; however, education regarding its role in perpetuating healthcare inequity remains limited in medical curricula. This study implemented and evaluated the impact of a student-led anti-racism programme on medical students’ perceptions of racial bias in medicine, awareness of, and confidence to advocate against racism in medicine. Method A total of 543 early stage medical students were invited to participate in the programme. Participants were assigned readings and videos exploring racial injustice in medicine and attended a virtual small-group discussion facilitated by faculty and students. Online surveys were used to collect pre- and post-programme data using Likert scales for response items. Open-ended questions were independently reviewed by three authors using reflexive thematic analysis. Results Sixty-three early-stage medical students enrolled in the programme, of which 42 completed the pre-programme survey. There was a 76% (n = 32) response rate for the post-programme survey. The majority of students (60%, n = 25) had no previous education about racism in medicine. From pre- to post-programme, there was a significant change in students’ perceived definition of race from genetic, biological, geographical, and cultural factors to socio-political factors (P < 0.0001). Significant increases in almost all factors assessing student awareness of racism and confidence to advocate against racism were observed. Student-identified barriers to discussing racism included lack of education and lived experience, fear of starting conflict and offending others. All survey respondents would recommend this programme to peers and 69% (n = 32) engaged in further topical self-directed education. Conclusion This simple and reproducible programme improved awareness and confidence to advocate against racism in medicine and resulted in a change in opinion regarding race-based medical practice. These findings are in line with best practice towards addressing racial bias in medicine, decolonizing medical curricula and strengthening anti-racism teaching of future physicians.

Is the use of the new Chronic Kidney Disease Epidemiology Consortium (CKD-EPI 2021) formula appropriate for the Spanish population?

Article

Feb 2023

Introduction: United States nephrology societies recommend changing from the CKD-EPI 2009 equation to the new CKD-EPI 2021 equation, which does not include the race coefficient, for calculating estimated glomerular filtration rate (eGFR). It is unknown how this change might affect the distribution of kidney disease in the predominantly Caucasian Spanish population. Methods: Two databases of adults from the province of Cádiz, DB-SIDICA (N = 264,217) and DB-PANDEMIA (N = 64,217), that had plasma creatinine measurements recorded between 2017 and 2021 were studied. Changes in eGFR and the consequent reclassification into different categories of the KDIGO 2012 classification resulting from substituting the CKD-EPI 2009 equation for the 2021 equation were calculated. Results: Compared to the 2009 equation, CKD-EPI 2021 yielded a higher eGFR, with a median of 3.8 mL/min/1.73 m2 (IQR 2.98-4.48) in DB-SIDICA and 3.89 mL/min/1.73 m2 (IQR 3.05-4.55) in DB-PANDEMIA. The first consequence was that 15.3% of the total population in DB-SIDICA and 15.1% of the total population in DB-PANDEMIA were reclassified into a higher category of eGFR, as were 28.1% and 27.3%, respectively, of the population with CKD (G3-G5); no subjects were classified into the more severe category. The second consequence was that the prevalence of kidney disease decreased from 9% to 7.5% in both cohorts. Conclusions: Implementing the CKD-EPI 2021 equation in the Spanish population, which is predominantly Caucasian, would increase eGFR by a modest amount (greater in men and those who are older or have a higher GFR). A significant proportion of the population would be classified into a higher eGFR category, with a consequent decrease in the prevalence of kidney disease.

Removal of the race coefficient and adjustment to individual BSA provide the most accurate estimation for GFR in Black adolescents

Preprint

Full-text available

Aug 2021

Introduction Obesity is more prevalent among minorities, increasing the risk for cardio-renal morbidity. We explored interactions between race, body mass index (BMI), and the risk of hyperfiltration associated with Obesity Related Glomerulopathy (ORG). Methods We created a cohort of women and girls ages 12-21 from the New York Metropolitan area using electronic health records. Glomerular filtration rate (GFR) was estimated in three ways: I) using the standard age recommended formulae, II) eGFRr – without a race-specific coefficient, and III) Absolute eGFRr – combing removal of the race coefficient and adjusting to individual body surface area. Multivariate logistic regression was used to analyze the relative contribution of risk factors for ORG associated hyperfiltration, defined by a threshold of ≥135ml/min/1.73m ² . Bland Altman analysis and Pearson’s coefficient assessed the correlation of each formula with creatinine clearance (CrCl). Results 7315 Black and 15,102 non-Black women and girls had simultaneous evaluation of kidney function and body measures. CrCl was available in 207 non-Black and 107 Black individuals as an internal validation. Simultaneous removal of the race coefficient and adjustment to individual BSA estimated GFR most accurately compared to CrCl, across BMI groups and between races. Hyperfiltration was more frequent in obese Black compared to non-Black individuals when using standard eGFR (20% vs. 6.5% respectively) but had a lower frequency after eliminating the race-specific coefficient (4.5% vs. 6.5%). Black race was independently associated with a higher risk of hyperfiltration with standard eGFR calculations (OR=3.43, 95% CI 2.95-3.99) and with lower risk when estimated by eGFRr (OR=0.56, 95% CI 0.45-0.70). Simultaneous removal of the race coefficient from GFR calculation and adjustment to individual BSA attenuated the difference in risk between races (OR=0.8, 95% CI 0.68-0.94). The combined correction agreed well with creatinine clearance (Pearson’s correlation coefficient r=0.64, 0.52 and 0.52 for absolute eGFRr, eGFR and eGFRr respectively. Conclusions Removal of the race coefficient from GFR estimating equations obscures obesity associated hyperfiltration among Black adolescents. This correction should be accompanied by adjustment to individual BSA to improve estimation of GFR to avoid misclassification of obesity related hyperfiltration.

Prospective cohort study to identify prevalence, risk factors and outcomes of infection associated kidney disease in a regional hospital in Malawi

Article

Full-text available

Nov 2022

Objectives Acute kidney injury (AKI) is a common and severe complication of community acquired infection, but data on impact in sub-Saharan Africa (SSA) are lacking. We determined prevalence, risk factors and outcomes of infection associated kidney disease in adults in Malawi. Design A prospective cohort study of adults admitted to hospital with infection, from February 2021 to June 2021, collecting demographic, clinical, laboratory and ultrasonography data. Setting Adults admitted to a regional hospital in Southern Region, Malawi. Primary and secondary outcome measures The primary outcomes were prevalence of kidney disease and mortality by Cox proportional hazard model. AKI was defined according to Kidney Disease Improving Global Outcomes (KDIGO) guidelines. Secondary outcomes were risk factors for AKI identified by logistic regression and prevalence of chronic kidney disease at 3 months. Results We recruited 101 patients presenting to hospital with infection. Median age was 38 years (IQR: 29–48 years), 88 had known HIV status of which 53 (60%) were living with HIV, and of these 42 (79%) were receiving antiretroviral therapy. AKI was present in 33/101 at baseline, of which 18/33 (55%) cases were severe (KDIGO stage 3). At 3 months, 28/94 (30%) participants had died, while 7/61 (11%) of survivors had chronic kidney disease. AKI was associated with older age (age: 60 years vs 40 years, OR: 3.88, 95% CI 1.82 to 16.64), and HIV positivity (OR: 4.08, 95% CI 1.28 to 15.67). Living with HIV was independently associated with death (HR: 3.97, 95% CI 1.07 to 14.69). Conclusions Kidney disease is common among hospitalised adults with infection in Malawi, with significant kidney impairment identified at 3 months. Our study highlights the difficulty in diagnosing acute and chronic kidney disease, and the need for more accurate methods than creatinine based estimated glomerular filtration rate (eGFR) equations for populations in Africa. Patients with kidney impairment identified in hospital should be prioritised for follow-up.

2009 CKD-EPI glomerular filtration rate estimation in Black individuals outside the United States: a systematic review and meta-analysis

Article

Full-text available

Oct 2022

Background The 2009 CKD-EPI is the most used equation to estimate glomerular filtration rate (GFR), with race being a factor thereof, increasing GFR by 16% in self-identified Black persons compared to non-Black persons. However, recent publications indicate that it might overestimate GFR for black adults outside the United States of America (USA). In this meta-analysis, we assessed the accuracy, evaluated by P30, of the 2009 CKD-EPI in estimating GFR with and without the race coefficient in black individuals outside the USA. Methods Searches on MEDLINE and EMBASE from inception to July 9th, 2022, with no language restriction, supplemented by manual reference searches. Studies that assessed the CKD-EPI P30 accuracy with or without the race coefficient in black adults outside the USA with an adequate method of GFR measurement were included. Data were extracted by independent pairs of reviewers and were pooled using a random-effects model. Results we included 11 studies, with a total of 1834 black adults from South America, Africa, and Europe. The race coefficient in the 2009 CKD-EPI equation significantly decreased P30 accuracy results (61.9% [95% CI, 53% to 70%] vs. 72.9% [95% CI, 66.7% to 78.3%]; p = 0.03). Conclusions Outside the USA, the 2009 CKD-EPI should not be used with the race coefficient, even though the 2009 CKD-EPI equation is not sufficiently accurate either way (<75%). Thus, we endorse KDIGO guidelines to use exogenous filtration markers when this may impact clinical conduct.

Designing Interventions Addressing Structural Racism to Reduce Kidney Health Disparities: A Report from an NIDDK Workshop

Article

Oct 2022

Structural racism embodies the many ways in which society fosters racial discrimination through "mutually reinforcing inequitable systems" that limit access to resources and opportunities that can promote health and well-being among marginalized communities. To achieve health equity, and kidney health equity more specifically, structural racism must be eliminated. In February 2022, the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) convened the "Designing Interventions that Address Structural Racism to Reduce Kidney Health Disparities" workshop which was aimed at describing the mechanisms through which structural racism contributes to health and healthcare disparities for people along the continuum of kidney disease; and identifying actionable opportunities for interventional research focused on dismantling or addressing the effects of structural racism. Participants identified six domains as key targets for interventions and future research: 1) apply an anti-racism lens, 2) promote structural interventions, 3) target multiple levels, 4) promote effective community and stakeholder engagement, 5) improve data collection, and 6) advance health equity through new healthcare models. There exists an urgent need for research to develop, implement and evaluate interventions that address the unjust systems, policies, and laws that generate and perpetuate inequities in kidney health.

Suboptimal monitoring and management in patients with unrecorded stage 3 chronic kidney disease in real-world settings: Insights from REVEAL-CKD

Article

Jul 2024
EUR J CLIN INVEST

Background Clinical practice guidelines for patients with chronic kidney disease (CKD) recommend regular monitoring and management of kidney function and CKD risk factors. However, the majority of patients with stage 3 CKD lack a diagnosis code, and data on the implementation of these recommendations in the real world are limited. Aim To assess the implementation of guideline‐directed monitoring and management practices in the real world in patients with stage 3 CKD without a recorded diagnosis code. Methods REVEAL‐CKD (NCT04847531) is a multinational, observational study of patients with stage 3 CKD. Eligible patients had ≥2 consecutive estimated glomerular filtration rate (eGFR) measurements indicative of stage 3 CKD recorded >90 and ≤730 days apart, lacked an International Classification of Diseases 9/10 diagnosis code corresponding to CKD any time before and up to 6 months after the second eGFR measurement. Testing of key measures of care quality were assessed. Results The study included 435,971 patients from 9 countries. In all countries, the prevalence of urinary albumin–creatinine ratio and albuminuria testing was low. Angiotensin‐converting enzyme inhibitor, angiotensin receptor blocker and statin prescriptions were highly variable, and sodium–glucose cotransporter‐2 inhibitor prescriptions remained below 21%. Blood pressure measurements were recorded in 20.2%–89.9% of patients. Conclusions Overall, a large proportion of patients with evidence of stage 3 CKD did not receive recommended, guideline‐directed monitoring and management. The variability in standard of care among countries demonstrates a clear opportunity to improve monitoring and management of these patients, most likely improving long‐term outcomes.

Consistency of metabolite associations with measured glomerular filtration rate in children and adults

Article

Full-text available

Apr 2024

Background There is interest in identifying novel filtration markers that lead to more accurate GFR estimates than current markers (creatinine and cystatin C) and are more consistent across demographic groups. We hypothesize that large-scale metabolomics can identify serum metabolites that are strongly influenced by glomerular filtration rate (GFR) and are more consistent across demographic variables than creatinine, which would be promising filtration markers for future investigation. Methods We evaluated the consistency of associations between measured GFR (mGFR) and 887 common, known metabolites quantified by an untargeted chromatography- and spectroscopy-based metabolomics platform (Metabolon) performed on frozen blood samples from 580 participants in Chronic Kidney Disease in Children (CKiD), 674 participants in Modification of Diet in Renal Disease (MDRD) Study and 962 participants in African American Study of Kidney Disease and Hypertension (AASK). We evaluated metabolite–mGFR correlation association with metabolite class, molecular weight, assay platform and measurement coefficient of variation (CV). Among metabolites with strong negative correlations with mGFR (r < −0.5), we assessed additional variation by age (height in children), sex, race and body mass index (BMI). Results A total of 561 metabolites (63%) were negatively correlated with mGFR. Correlations with mGFR were highly consistent across study, sex, race and BMI categories (correlation of metabolite–mGFR correlations between 0.88 and 0.95). Amino acids, carbohydrates and nucleotides were more often negatively correlated with mGFR compared with lipids, but there was no association with metabolite molecular weight, liquid chromatography/mass spectrometry platform and measurement CV. Among 114 metabolites with strong negative associations with mGFR (r < −0.5), 27 were consistently not associated with age (height in children), sex or race. Conclusions The majority of metabolite–mGFR correlations were negative and consistent across sex, race, BMI and study. Metabolites with consistent strong negative correlations with mGFR and non-association with demographic variables may represent candidate markers to improve estimation of GFR.

Expanding the access to kidney transplantation: Strategies for kidney transplant programs

Article

Apr 2024

Kidney transplantation is the most successful kidney replacement therapy available, resulting in improved recipient survival and societal cost savings. Yet, nearly 70 years after the first successful kidney transplant, there are still numerous barriers and untapped opportunities that constrain the access to transplant. The literature describing these barriers is extensive, but the practices and processes to solve them are less clear. Solutions must be multidisciplinary and be the product of strong partnerships among patients, their networks, health care providers, and transplant programs. Transparency in the referral, evaluation, and listing process as well as organ selection are paramount to build such partnerships. Providing early culturally congruent and patient‐centered education as well as maximizing the use of local resources to facilitate the transplant work up should be prioritized. Every opportunity to facilitate pre‐emptive kidney transplantation and living donation must be taken. Promoting the use of telemedicine and kidney paired donation as standards of care can positively impact the work up completion and maximize the chances of a living donor kidney transplant.

Evaluating and Improving the Performance and Racial Fairness of Algorithms for GFR Estimation

Conference Paper

Feb 2024

Demographic bias in misdiagnosis by computational pathology models

Article

Full-text available

Apr 2024
NAT MED

Despite increasing numbers of regulatory approvals, deep learning-based computational pathology systems often overlook the impact of demographic factors on performance, potentially leading to biases. This concern is all the more important as computational pathology has leveraged large public datasets that underrepresent certain demographic groups. Using publicly available data from The Cancer Genome Atlas and the EBRAINS brain tumor atlas, as well as internal patient data, we show that whole-slide image classification models display marked performance disparities across different demographic groups when used to subtype breast and lung carcinomas and to predict IDH1 mutations in gliomas. For example, when using common modeling approaches, we observed performance gaps (in area under the receiver operating characteristic curve) between white and Black patients of 3.0% for breast cancer subtyping, 10.9% for lung cancer subtyping and 16.0% for IDH1 mutation prediction in gliomas. We found that richer feature representations obtained from self-supervised vision foundation models reduce performance variations between groups. These representations provide improvements upon weaker models even when those weaker models are combined with state-of-the-art bias mitigation strategies and modeling choices. Nevertheless, self-supervised vision foundation models do not fully eliminate these discrepancies, highlighting the continuing need for bias mitigation efforts in computational pathology. Finally, we demonstrate that our results extend to other demographic factors beyond patient race. Given these findings, we encourage regulatory and policy agencies to integrate demographic-stratified evaluation into their assessment guidelines.

Renal Handling of Lithium

Chapter

Feb 2024

In a 1989 review entitled “Long-term treatment with lithium and renal function: A review and reappraisal,” the pioneering Danish psychopharmacologist Mogens Schou concluded: “The fear of eventual kidney insufficiency as a result of long-term lithium treatment can be set at rest” [1]. Despite the certainty advanced by the preeminent authority on lithium at that time – the man responsible for all of the early data on lithium’s efficacy and the first double-blind placebo-controlled trials – fear of lithium’s long-term renal adverse effects remains a significant concern to clinicians, a concern that is often disproportionate to the emerging data in this area.

A Canadian Commentary on the NKF-ASN Task Force Recommendations on Reassessing the Inclusion of Race in Diagnosing Kidney Disease

Article

Full-text available

Nov 2023

In 2021, a committee was commissioned by the Canadian Society of Nephrology to comment on the 2021 National Kidney Foundation–American Society of Nephrology Task Force recommendations on the use of race in glomerular filtration rate estimating equations. The committee met on numerous occasions and agreed on several recommendations. However, the committee did not achieve unanimity, with a minority group disagreeing with the scope of the commentary. As a result, this report presents the viewpoint of the majority members. We endorsed many of the recommendations from the National Kidney Foundation–American Society of Nephrology Task Force, most importantly that race should be removed from the estimated glomerular filtration rate creatinine-based equation. We recommend an immediate implementation of the new Chronic Kidney Disease Epidemiology Collaboration equation (2021), which does not discriminate among any group while maintaining precision. Additionally, we recommend that Canadian laboratories and provincial kidney organizations advocate for increased testing and access to cystatin C because the combination of cystatin C and creatinine in revised equations leads to more precise estimates. Finally, we recommend that future research studies evaluating the implementation of the new equations and changes to screening, diagnosis, and management across provincial health programs be prioritized in Canada.

ELIMINATING THE RACE COEFFICIENT IN KIDNEY FUNCTION ESTIMATING EQUATIONS: THE CENTER DID HOLD

Article

Jan 2023
Trans Am Clin Climatol Assoc

Chi-yuan Hsu

Performance of Future Glomerular Filtration Rate (fGFR) Equation by Race in a Large, Racially Diverse Patient Cohort Undergoing Nephrectomy for Renal Cell Carcinoma

Article

Oct 2023
UROLOGY

Historic use race-based spirometry values lowered transplant priority for Black patients

Article

Oct 2023
CHEST

Replacing Race in Clinical Algorithms: The Need for Thoughtful Evaluations

Article

Sep 2023
J Pediatr

Sara Khor

African American/black race, Apolipoprotein L1, and the indirect glomerular filtration measure serum creatinine among persons living with HIV

Article

Aug 2023
AIDS

Objective: Accurate estimation of kidney function is critical among people living with HIV (PLWH) to avoid under-dosing of antiretroviral therapies and ensure timely referral for kidney transplantation. Existing estimation equations for kidney function include race, the appropriateness of which has been debated. Given advancements in understanding of race and the necessity of accuracy in kidney function estimation, this study aimed to examine whether race, or genetic factors, improved prediction of serum creatinine among PLWH. Design: This cross-sectional study utilized data from the Center for AIDS Research Network of Integrated Clinical Systems cohort (2008-2018). The outcome was baseline serum creatinine. Methods: Ordinary least squares regression was used to examine whether inclusion of race or genetic factors (apolipoprotein-L1 (APOL1) variants and genetic African ancestry) improved serum creatinine prediction. A reduction in root mean squared error (RMSE) >2% was a clinically relevant improvement in predictive ability. Results: There were 4,183 PLWH included. Among PLWH whose serum creatinine <1.7 mg/dL, race was significantly associated with serum creatinine (β = 0.06, SE = 0.01, p < 0.001), but did not improve predictive ability. African ancestry and APOL1 variants similarly failed to improve predictive ability. Whereas, when serum creatinine ≥1.7 mg/dL, inclusion of race reduced the RMSE by 2.1%, indicating improvement in predictive ability. APOL1 variants further improved predictive ability by reducing the RMSE by 2.9. Conclusions: These data suggest that, among PLWH, inclusion of race or genetic factors may only be warranted at higher serum creatinine levels. Work eliminating existing healthcare disparities while preserving the utility of estimating equations is needed.

Fair admission risk prediction with proportional multicalibration

Article

Jan 2023

Fair calibration is a widely desirable fairness criteria in risk prediction contexts. One way to measure and achieve fair calibration is with multicalibration. Multicalibration constrains calibration error among flexibly-defined subpopulations while maintaining overall calibration. However, multicalibrated models can exhibit a higher percent calibration error among groups with lower base rates than groups with higher base rates. As a result, it is possible for a decision-maker to learn to trust or distrust model predictions for specific groups. To alleviate this, we propose proportional multicalibration, a criteria that constrains the percent calibration error among groups and within prediction bins. We prove that satisfying proportional multicalibration bounds a model's multicalibration as well its differential calibration, a fairness criteria that directly measures how closely a model approximates sufficiency. Therefore, proportionally calibrated models limit the ability of decision makers to distinguish between model performance on different patient groups, which may make the models more trustworthy in practice. We provide an efficient algorithm for post-processing risk prediction models for proportional multicalibration and evaluate it empirically. We conduct simulation studies and investigate a real-world application of PMC-postprocessing to prediction of emergency department patient admissions. We observe that proportional multicalibration is a promising criteria for controlling simultaneous measures of calibration fairness of a model over intersectional groups with virtually no cost in terms of classification performance.

Pharmacogenetics: Using science to reduce racial bias in pain management?

Article

Aug 2023
J AM GERIATR SOC

A Call to Comprehensively Understand Our Patients to Provide Equitable Pediatric Urological Care

Article

Jun 2023

Algorithmic fairness in artificial intelligence for medicine and healthcare

Article

Jun 2023
Nat. Biomed. Eng.

In healthcare, the development and deployment of insufficiently fair systems of artificial intelligence (AI) can undermine the delivery of equitable care. Assessments of AI models stratified across subpopulations have revealed inequalities in how patients are diagnosed, treated and billed. In this Perspective, we outline fairness in machine learning through the lens of healthcare, and discuss how algorithmic biases (in data acquisition, genetic variation and intra-observer labelling variability, in particular) arise in clinical workflows and the resulting healthcare disparities. We also review emerging technology for mitigating biases via disentanglement, federated learning and model explainability, and their role in the development of AI-based software as a medical device.

Reevaluating the Role of Race and Ethnicity in Diabetes Screening

Preprint

Full-text available

Jun 2023

There is active debate over whether to consider patient race and ethnicity when estimating disease risk. By accounting for race and ethnicity, it is possible to improve the accuracy of risk predictions, but there is concern that their use may encourage a racialized view of medicine. In diabetes risk models, despite substantial gains in statistical accuracy from using race and ethnicity, the gains in clinical utility are surprisingly modest. These modest clinical gains stem from two empirical patterns: first, the vast majority of individuals receive the same screening recommendation regardless of whether race or ethnicity are included in risk models; and second, for those who do receive different screening recommendations, the difference in utility between screening and not screening is relatively small. Our results are based on broad statistical principles, and so are likely to generalize to many other risk-based clinical decisions.

Racial and Ethnic Bias in Risk Prediction Models for Colorectal Cancer Recurrence When Race and Ethnicity Are Omitted as Predictors

Article

Full-text available

Jun 2023

Importance: Including race and ethnicity as a predictor in clinical risk prediction algorithms has received increased scrutiny, but there continues to be a lack of empirical studies addressing whether simply omitting race and ethnicity from the algorithms will ultimately affect decision-making for patients of minoritized racial and ethnic groups. Objective: To examine whether including race and ethnicity as a predictor in a colorectal cancer recurrence risk algorithm is associated with racial bias, defined as racial and ethnic differences in model accuracy that could potentially lead to unequal treatment. Design, setting, and participants: This retrospective prognostic study was conducted using data from a large integrated health care system in Southern California for patients with colorectal cancer who received primary treatment between 2008 and 2013 and follow-up until December 31, 2018. Data were analyzed from January 2021 to June 2022. Main outcomes and measures: Four Cox proportional hazards regression prediction models were fitted to predict time from surveillance start to cancer recurrence: (1) a race-neutral model that explicitly excluded race and ethnicity as a predictor, (2) a race-sensitive model that included race and ethnicity, (3) a model with 2-way interactions between clinical predictors and race and ethnicity, and (4) separate models by race and ethnicity. Algorithmic fairness was assessed using model calibration, discriminative ability, false-positive and false-negative rates, positive predictive value (PPV), and negative predictive value (NPV). Results: The study cohort included 4230 patients (mean [SD] age, 65.3 [12.5] years; 2034 [48.1%] female; 490 [11.6%] Asian, Hawaiian, or Pacific Islander; 554 [13.1%] Black or African American; 937 [22.1%] Hispanic; and 2249 [53.1%] non-Hispanic White). The race-neutral model had worse calibration, NPV, and false-negative rates among racial and ethnic minority subgroups than non-Hispanic White individuals (eg, false-negative rate for Hispanic patients: 12.0% [95% CI, 6.0%-18.6%]; for non-Hispanic White patients: 3.1% [95% CI, 0.8%-6.2%]). Adding race and ethnicity as a predictor improved algorithmic fairness in calibration slope, discriminative ability, PPV, and false-negative rates (eg, false-negative rate for Hispanic patients: 9.2% [95% CI, 3.9%-14.9%]; for non-Hispanic White patients: 7.9% [95% CI, 4.3%-11.9%]). Inclusion of race interaction terms or using race-stratified models did not improve model fairness, likely due to small sample sizes in subgroups. Conclusions and relevance: In this prognostic study of the racial bias in a cancer recurrence risk algorithm, removing race and ethnicity as a predictor worsened algorithmic fairness in multiple measures, which could lead to inappropriate care recommendations for patients who belong to minoritized racial and ethnic groups. Clinical algorithm development should include evaluation of fairness criteria to understand the potential consequences of removing race and ethnicity for health inequities.

Estimating baseline creatinine to detect acute kidney injury in patients with chronic kidney disease

Article

Jun 2023
NEPHROLOGY

Background: Accurately estimating baseline kidney function is essential for diagnosing acute kidney injury (AKI) in patients with chronic kidney disease (CKD). We developed and evaluated novel equations to estimate baseline creatinine in patients with AKI on CKD. Methods: We retrospectively analysed 5649 adults with AKI out of 11 254 CKD patients, dividing them evenly into derivation and validation groups. Using quantiles regression, we created equations to estimate baseline creatinine, considering historical creatinine values, months since measurement, age, and sex from the derivation dataset. We assessed performance against back-estimation equations and unadjusted historical creatinine values using the validation dataset. Results: The optimal equation adjusted the most recent creatinine value for time since measurement and sex. Estimates closely matched the actual baseline at AKI onset, with median (95% confidence interval) differences of just 0.9% (-0.8% to 2.1%) and 0.6% (-1.6% to 3.9%) when the most recent value was within 6 months to 30 days and 2 years to 6 months before AKI onset, respectively. The equation improved AKI event reclassification by an additional 2.5% (2.0% to 3.0%) compared to the unadjusted most recent creatinine value and 7.3% (6.2% to 8.4%) compared to the CKD-EPI 2021 back-estimation equation. Conclusion: Creatinine levels drift in patients with CKD, causing false positives in AKI detection without adjustment. Our novel equation adjusts the most recent creatinine value for drift over time. It provides more accurate baseline creatinine estimation in patients with suspected AKI on CKD, which reduces false-positive AKI detection, improving patient care and management.

Awareness of Racial and Ethnic Bias and Potential Solutions to Address Bias With Use of Health Care Algorithms

Article

Full-text available

Jun 2023

Importance: Algorithms are commonly incorporated into health care decision tools used by health systems and payers and thus affect quality of care, access, and health outcomes. Some algorithms include a patient's race or ethnicity among their inputs and can lead clinicians and decision-makers to make choices that vary by race and potentially affect inequities. Objective: To inform an evidence review on the use of race- and ethnicity-based algorithms in health care by gathering public and stakeholder perspectives about the repercussions of and efforts to address algorithm-related bias. Design, setting, and participants: Qualitative methods were used to analyze responses. Responses were initially open coded and then consolidated to create a codebook, with themes and subthemes identified and finalized by consensus. This qualitative study was conducted from May 4, 2021, through December 7, 2022. Forty-two organization representatives (eg, clinical professional societies, universities, government agencies, payers, and health technology organizations) and individuals responded to the request for information. Main outcomes and measures: Identification of algorithms with the potential for race- and ethnicity-based biases and qualitative themes. Results: Forty-two respondents identified 18 algorithms currently in use with the potential for bias, including, for example, the Simple Calculated Osteoporosis Risk Estimation risk prediction tool and the risk calculator for vaginal birth after cesarean section. The 7 qualitative themes, with 31 subthemes, included the following: (1) algorithms are in widespread use and have significant repercussions, (2) bias can result from algorithms whether or not they explicitly include race, (3) clinicians and patients are often unaware of the use of algorithms and potential for bias, (4) race is a social construct used as a proxy for clinical variables, (5) there is a lack of standardization in how race and social determinants of health are collected and defined, (6) bias can be introduced at all stages of algorithm development, and (7) algorithms should be discussed as part of shared decision-making between the patient and clinician. Conclusions and relevance: This qualitative study found that participants perceived widespread and increasing use of algorithms in health care and lack of oversight, potentially exacerbating racial and ethnic inequities. Increasing awareness for clinicians and patients and standardized, transparent approaches for algorithm development and implementation may be needed to address racial and ethnic biases related to algorithms.

Prevalence of undiagnosed stage 3 chronic kidney disease in France, Germany, Italy, Japan and the USA: results from the multinational observational REVEAL-CKD study

Article

Full-text available

May 2023

Objectives REVEAL-CKD aims to estimate the prevalence of, and factors associated with, undiagnosed stage 3 chronic kidney disease (CKD). Design Multinational, observational study. Setting Data from six country-specific electronic medical records and/or insurance claims databases from five countries (France, Germany, Italy, Japan and the USA [two databases]). Participants Eligible participants (≥18 years old) had ≥2 consecutive estimated glomerular filtration rate (eGFR) measurements (calculated from serum creatinine values, sex and age) taken from 2015 onwards that were indicative of stage 3 CKD (≥30 and <60 mL/min/1.73 m ² ). Undiagnosed cases lacked an International Classification of Diseases 9/10 diagnosis code for CKD (any stage) any time before, and up to 6 months after, the second qualifying eGFR measurement (study index). Main outcome measures The primary outcome was point prevalence of undiagnosed stage 3 CKD. Time to diagnosis was assessed using the Kaplan-Meier approach. Factors associated with lacking a CKD diagnosis and risk of diagnostic delay were assessed using logistic regression adjusted for baseline covariates. Results The prevalence of undiagnosed stage 3 CKD was 95.5% (19 120/20 012 patients) in France, 84.3% (22 557/26 767) in Germany, 77.0% (50 547/65 676) in Italy, 92.1% (83 693/90 902) in Japan, 61.6% (13 845/22 470) in the US Explorys Linked Claims and Electronic Medical Records Data database and 64.3% (161 254/250 879) in the US TriNetX database. The prevalence of undiagnosed CKD increased with age. Factors associated with undiagnosed CKD were female sex (vs male, range of odds ratios across countries: 1.29–1.77), stage 3a CKD (vs 3b, 1.81–3.66), no medical history (vs a history) of diabetes (1.26–2.77) or hypertension (1.35–1.78). Conclusions There are substantial opportunities to improve stage 3 CKD diagnosis, particularly in female patients and older patients. The low diagnosis rates in patients with comorbidities that put them at risk of disease progression and complications require attention. Trial registration NCT04847531 .

Biases in Machine Learning in Healthcare

Chapter

Feb 2023

Machine learning in healthcare (MLHC) has the potential to revolutionize healthcare and health systems research. However, these benefits must be weighed against the risks of MLHC in perpetuating or even magnifying existing health disparities. This chapter discusses existing and historical biases in clinical medicine, examines the potential hazards associated with MLHC implementation, and considers possible solutions to mitigate these concerns.

Impacts of removing race from the calculation of the Kidney Donor Profile Index

Article

Jan 2023
AM J TRANSPLANT

The kidney donor risk index (KDRI), standardized as the kidney donor profile index (KDPI), estimates graft failure risk for organ allocation and includes a coefficient for the Black donor race that could create disparities. This study used the Scientific Registry of Transplant Recipients data to recalculate KDRI coefficients with and without the Black race variable for deceased donor kidney transplants from 1995 to 2005 (n = 69 244). The recalculated coefficients were applied to deceased kidney donors from 2015 to 2021 (n = 72 926) to calculate KDPI. Removing the Black race variable had a negligible impact on the model's predictive ability. When the Black race variable was removed, the proportion of Black donors above KDPI 85%, a category with a higher risk of organ nonuse, declined from 31.09% to 17.75%, closer to the 15.68% above KDPI 85% among non-Black donors. KDPI represents percentiles relative to all other donors, so the number of Black donors moving below KDPI 86% was roughly equal to the number of non-Black donors moving above KDPI 85%. Removing the Black donor indicator from KDRI/KDPI may improve equity without substantial overall impact on the transplantation system, though further improvement may require the use of absolute measures of donor risk KDRI rather than relative measures of risk KDPI.

Health Services Research as a Vehicle for Informed Policy Advocacy

Article

Nov 2022

Race and the False Precision of Glomerular Filtration Rate Estimates

Article

Aug 2020

Ashwini R Sehgal

Hidden in Plain Sight — Reconsidering the Use of Race Correction in Clinical Algorithms

Article

Jun 2020

Physicians still lack consensus on the meaning of race. When the Journal took up the topic in 2003 with a debate about the role of race in medicine, one side argued that racial and ethnic categories reflected underlying population genetics and could be clinically useful.1 Others held that any small benefit was outweighed by potential harms that arose from the long, rotten history of racism in medicine.2 Weighing the two sides, the accompanying Perspective article concluded that though the concept of race was "fraught with sensitivities and fueled by past abuses and the potential for future abuses," race-based medicine still had potential: "it seems unwise to abandon the practice of recording race when we have barely begun to understand the architecture of the human genome."3 The next year, a randomized trial showed that a combination of hydralazine and isosorbide dinitrate reduced mortality due to heart failure among patients who identified themselves as black. The Food and Drug Administration granted a race-specific indication for that product, BiDil, in 2005.4 Even though BiDil's ultimate commercial failure cast doubt on race-based medicine, it did not lay the approach to rest. Prominent geneticists have repeatedly called on physicians to take race seriously,5,6 while distinguished social scientists vehemently contest these calls.7,8 Our understanding of race and human genetics has advanced considerably since 2003, yet these insights have not led to clear guidelines on the use of race in medicine. The result is ongoing conflict between the latest insights from population genetics and the clinical implementation of race. For example, despite mounting evidence that race is not a reliable proxy for genetic difference, the belief that it is has become embedded, sometimes insidiously, within medical practice. One subtle insertion of race into medicine involves diagnostic algorithms and practice guidelines that adjust or "correct" their outputs on the basis of a patient's race or ethnicity. Physicians use these algorithms to individualize risk assessment and guide clinical decisions. By embedding race into the basic data and decisions of health care, these algorithms propagate race-based medicine. Many of these race-adjusted algorithms guide decisions in ways that may direct more attention or resources to white patients than to members of racial and ethnic minorities. To illustrate the potential dangers of such practices, we have compiled a partial list of raceadjusted algorithms (Table 1). We explore several of them in detail here. Given their potential to perpetuate or even amplify race-based health inequities, they merit thorough scrutiny.

Kidney Disease, Race, and GFR Estimation

Article

May 2020

Assessment of GFR is central to clinical practice, research, and public health. Current Kidney Disease Improving Global Outcomes guidelines recommend measurement of serum creatinine to estimate GFR as the initial step in GFR evaluation. Serum creatinine is influenced by creatinine metabolism as well as GFR; hence, all equations to estimate GFR from serum creatinine include surrogates for muscle mass, such as age, sex, race, height, or weight. The guideline-recommended equation in adults (the 2009 Chronic Kidney Disease Epidemiology Collaboration creatinine equation) includes a term for race (specified as black versus nonblack), which improves the accuracy of GFR estimation by accounting for differences in non-GFR determinants of serum creatinine by race in the study populations used to develop the equation. In that study, blacks had a 16% higher average measured GFR compared with nonblacks with the same age, sex, and serum creatinine. The reasons for this difference are only partly understood, and the use of race in GFR estimation has limitations. Some have proposed eliminating the race coefficient, but this would induce a systematic underestimation of measured GFR in blacks, with potential unintended consequences at the individual and population levels. We propose a more cautious approach that maintains and improves accuracy of GFR estimates and avoids disadvantaging any racial group. We suggest full disclosure of use of race in GFR estimation, accommodation of those who decline to identify their race, and shared decision making between health care providers and patients. We also suggest mindful use of cystatin C as a confirmatory test as well as clearance measurements. It would be preferable to avoid specification of race in GFR estimation if there was a superior, evidence-based substitute. The goal of future research should be to develop more accurate methods for GFR estimation that do not require use of race or other demographic characteristics.

Reconsidering the Consequences of Using Race to Estimate Kidney Function

Article

Jun 2019
J Am Med Assoc

Clinicians estimate kidney function to guide important medical decisions across a wide range of settings, including assessing the safety of radiology studies, choosing chemotherapy, and reviewing the use of common nonprescription medications such as nonsteroidal anti-inflammatory drugs. Because direct measurement of kidney function is infeasible at the bedside, the usual approach involves using estimating equations that rely on serum creatinine. These equations assign a higher estimated glomerular filtration rate (eGFR) to patients who are identified as black. Yet in some medical and social science disciplines, a consensus has emerged that race is a social construct rather than a biological one.¹ In this Viewpoint, we argue that the use of kidney function estimating equations that include race as a variable cause problems for transparency and unduly restrict access to care in some cases, yet offer only modest benefits to precision.

Summary of Recommendation Statements

Article

Jan 2013

The Kidney Disease: Improving Global Outcomes (KDIGO) 2012 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease (CKD) serves to update the 2002 KDOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification following a decade of focused research and clinical practice in CKD. The document aims to provide state-of-the-art guidance on the evaluation, management and treatment for all patients with CKD. Specifically, the guideline retains the definition of CKD but presents an enhanced classification framework for CKD; elaborates on the identification and prognosis of CKD; discusses the management of progression and complications of CKD; and expands on the continuum of CKD care: timing of specialist referral, ongoing management of people with progressive CKD, timing of the initiation of dialysis, and finally the implementation of a treatment program which includes comprehensive conservative management. The development of the guideline followed an explicit process of evidence review and appraisal. Treatment approaches are addressed in each chapter and guideline recommendations are based on systematic reviews of relevant trials. Practical comments or statements which serve as educational purposes are ungraded, but included as important information for the readership. Appraisal of the quality of the evidence and the strength of recommendations followed the GRADE approach. Ongoing areas of controversies, limitations of the evidence, and international relevance are discussed and additional suggestions are provided for future research.

KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease: summary of recommendation statements

Jan 2013
KIDNEY INT
5-14

Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease: summary of recommendation statements. Kidney Int. 2013;3(suppl 1):5-14. doi: doi:10.1038/kisup.2012.77

Clinical Implications of Removing Race From Estimates of Kidney Function

Abstract

No full-text available

Recommended publications

New Practical Approachs to Estimation of Glomerular Filtration Rate in Adult: A Review