Try a new search

Format these results:

Searched for:

person:ww857

in-biosketch:yes

Total Results:

17


Analysis of hospital readmissions with competing risks

Wu, Wenbo; He, Kevin; Shi, Xu; Schaubel, Douglas E; Kalbfleisch, John D
The 30-day hospital readmission rate has been used in provider profiling for evaluating inter-provider care coordination, medical cost effectiveness, and patient quality of life. Current profiling analyzes use logistic regression to model 30-day readmission as a binary outcome, but one disadvantage of this approach is that this outcome is strongly affected by competing risks (e.g., death). Thus, one, perhaps unintended, consequence is that if two facilities have the same rates of readmission, the one with the higher rate of competing risks will have the lower 30-day readmission rate. We propose a discrete time competing risk model wherein the cause-specific readmission hazard is used to assess provider-level effects. This approach takes account of the timing of events and focuses on the readmission rates which are of primary interest. The quality measure, then is a standardized readmission ratio, akin to a standardized mortality ratio. This measure is not systematically affected by the rate of competing risks. To facilitate the estimation and inference of a large number of provider effects, we develop an efficient Blockwise Inversion Newton algorithm, and a stabilized robust score test that overcomes the conservative nature of the classical robust score test. An application to dialysis patients demonstrates improved profiling, model fitting, and outlier detection over existing methods.
PMID: 35899312
ISSN: 1477-0334
CID: 5273972

Improving large-scale estimation and inference for profiling health care providers

Wu, Wenbo; Yang, Yuan; Kang, Jian; He, Kevin
Provider profiling has been recognized as a useful tool in monitoring health care quality, facilitating inter-provider care coordination, and improving medical cost-effectiveness. Existing methods often use generalized linear models with fixed provider effects, especially when profiling dialysis facilities. As the number of providers under evaluation escalates, the computational burden becomes formidable even for specially designed workstations. To address this challenge, we introduce a serial blockwise inversion Newton algorithm exploiting the block structure of the information matrix. A shared-memory divide-and-conquer algorithm is proposed to further boost computational efficiency. In addition to the computational challenge, the current literature lacks an appropriate inferential approach to detecting providers with outlying performance especially when small providers with extreme outcomes are present. In this context, traditional score and Wald tests relying on large-sample distributions of the test statistics lead to inaccurate approximations of the small-sample properties. In light of the inferential issue, we develop an exact test of provider effects using exact finite-sample distributions, with the Poisson-binomial distribution as a special case when the outcome is binary. Simulation analyses demonstrate improved estimation and inference over existing methods. The proposed methods are applied to profiling dialysis facilities based on emergency department encounters using a dialysis patient database from the Centers for Medicare & Medicaid Services.
PMID: 35318706
ISSN: 1097-0258
CID: 5228192

Scalable proximal methods for cause-specific hazard modeling with time-varying coefficients

Wu, Wenbo; Taylor, Jeremy M G; Brouwer, Andrew F; Luo, Lingfeng; Kang, Jian; Jiang, Hui; He, Kevin
Survival modeling with time-varying coefficients has proven useful in analyzing time-to-event data with one or more distinct failure types. When studying the cause-specific etiology of breast and prostate cancers using the large-scale data from the Surveillance, Epidemiology, and End Results (SEER) Program, we encountered two major challenges that existing methods for estimating time-varying coefficients cannot tackle. First, these methods, dependent on expanding the original data in a repeated measurement format, result in formidable time and memory consumption as the sample size escalates to over one million. In this case, even a well-configured workstation cannot accommodate their implementations. Second, when the large-scale data under analysis include binary predictors with near-zero variance (e.g., only 0.6% of patients in our SEER prostate cancer data had tumors regional to the lymph nodes), existing methods suffer from numerical instability due to ill-conditioned second-order information. The estimation accuracy deteriorates further with multiple competing risks. To address these issues, we propose a proximal Newton algorithm with a shared-memory parallelization scheme and tests of significance and nonproportionality for the time-varying effects. A simulation study shows that our scalable approach reduces the time and memory costs by orders of magnitude and enjoys improved estimation accuracy compared with alternative approaches. Applications to the SEER cancer data demonstrate the real-world performance of the proximal Newton algorithm.
PMID: 35092553
ISSN: 1572-9249
CID: 5228162

Test-specific funnel plots for healthcare provider profiling leveraging individual- and summary-level information

Wu, Wenbo; Kuriakose, Jonathan P; Weng, Wenjiang; Burney, Richard E; He, Kevin
ORIGINAL:0015694
ISSN: 1387-3741
CID: 5273992

The Impact of COVID-19 on Post-Discharge Outcomes for Dialysis Patients in the United States: Evidence from Medicare Claims Data

Wu, Wenbo; Gremel, Garrett W; He, Kevin; Messanaa, Joseph M; Sen, Ananda; Segal, Jonathan H; Dahlerus, Claudia; Hirth, Richard A; Kang, Jian; Wisniewski, Karen; Nahra, Tammie; Padilla, Robin; Tong, Lan; Gu, Haoyu; Wang, Xi; Slowey, Megan; Eckard, Ashley; Ding, Xuemei; Borowicz, Lisa; Du, Juan; Frye, Brandon; Kalbfleisch, John D
ORIGINAL:0015568
ISSN: 2641-7650
CID: 5228462

Acute Kidney Injury Requiring Dialysis and Incident Dialysis Patient Outcomes in US Outpatient Dialysis Facilities

Dahlerus, Claudia; Segal, Jonathan H; He, Kevin; Wu, Wenbo; Chen, Shu; Shearon, Tempie H; Sun, Yating; Pearson, Aaron; Li, Xiang; Messana, Joseph M
BACKGROUND AND OBJECTIVES:About 30% of patients with AKI may require ongoing dialysis in the outpatient setting after hospital discharge. A 2017 Centers for Medicare & Medicaid Services policy change allows Medicare beneficiaries with AKI requiring dialysis to receive outpatient treatment in dialysis facilities. Outcomes for these patients have not been reported. We compare patient characteristics and mortality among patients with AKI requiring dialysis and patients without AKI requiring incident dialysis. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS:We used a retrospective cohort design with 2017 Medicare claims to follow outpatients with AKI requiring dialysis and patients without AKI requiring incident dialysis up to 365 days. Outcomes are unadjusted and adjusted mortality using Kaplan-Meier estimation for unadjusted survival probability, Poisson regression for monthly mortality, and Cox proportional hazards modeling for adjusted mortality. RESULTS:<0.001), which persisted through month 7. Overall adjusted mortality risk was 22% higher for patients with AKI requiring dialysis (1.22; 95% confidence interval, 1.17 to 1.27). CONCLUSIONS:In fully adjusted analyses, patients with AKI requiring dialysis had higher early mortality compared with patients without AKI requiring incident dialysis, but these differences declined after several months. Differences were also observed by age, race, and ethnicity within both patient cohorts.
PMID: 34045300
ISSN: 1555-905x
CID: 5228182

Surgeon Characteristics and Dialysis Vascular Access Outcomes in the United States: A Retrospective Cohort Study

Shahinian, Vahakn B; Zhang, Xiaosong; Tilea, Anca M; He, Kevin; Schaubel, Douglas E; Wu, Wenbo; Pisoni, Ronald; Robinson, Bruce; Saran, Rajiv; Woodside, Kenneth J
RATIONALE & OBJECTIVE:An arteriovenous fistula (AVF) is the preferred access for most patients receiving maintenance hemodialysis, but maturation failure remains a challenge. Surgeon characteristics have been proposed as contributors to AVF success. We examined variation in AVF placement and AVF outcomes by surgeon and surgeon characteristics. STUDY DESIGN:Retrospective cohort study. SETTING & PARTICIPANTS:National Medicare claims and web-based data submitted by dialysis facilities on maintenance hemodialysis patients from 2009 through 2015. EXPOSURES:Patient characteristics, including demographics and comorbid conditions; surgeon characteristics, including specialty, prior volume of AVF placements, and years since medical school graduation. OUTCOMES:Percent of access placements that were an AVF from 2009 to 2015 (designated AVF placement), and percent of AVFs with successful use within 6 months of placement (maturation) from 2013 to 2014. ANALYTICAL APPROACH:Multilevel logistic regression models examining the association of surgeon characteristics with the outcomes, adjusted for patient characteristics and dialysis facilities as random effects. RESULTS:Among 4,959 surgeons placing 467,827 accesses, median AVF placement was 71% (IQR, 59%-84%). More recent year of medical school graduation and general surgery specialty (vs vascular, cardiothoracic, or transplantation surgery) were associated with higher odds of AVF placement. Among 2,770 surgeons placing 49,826 AVFs, the median AVF maturation rate was 59% (IQR, 44%-71%). More recent year of medical school graduation, but not surgical specialty, was associated with higher odds of AVF maturation. Greater prior volume of AVF placement was associated with higher odds of AVF maturation: OR of 1.46 (95% CI, 1.37-1.57) for highest (>84 AVF placements in 2years) versus lowest (<14) volume quintile. LIMITATIONS:The study relied on administrative data, limiting capture of some factors affecting access outcomes. CONCLUSIONS:There is substantial surgeon-level variation in AVF placements and AVF maturation. Surgeons' prior volume of AVF placements is strongly associated with AVF maturation.
PMID: 31585684
ISSN: 1523-6838
CID: 5228172

Organic Pollutant Exposure and CKD: A Chronic Renal Insufficiency Cohort Pilot Study

Charytan, David M; Wu, Wenbo; Liu, Mengling; Li, Zhong-Min; Kannan, Kurunthachalam; Trasande, Leonardo; Pal, Vineet Kumar; Lee, Sunmi; Trachtman, Howard; Appel, Lawrence J.; Chen, Jing; Cohen, Debbie L.; Feldman, Harold I.; Go, Alan S.; Lash, James P.; Nelson, Robert G.; Rahman, Mahboob; Rao, Panduranga S.; Shah, Vallabh O; Unruh, Mark L
ORIGINAL:0017117
ISSN: 2590-0595
CID: 5634782

Neighborhood Segregation and Access to Live Donor Kidney Transplantation

Li, Yiting; Menon, Gayathri; Kim, Byoungjun; Bae, Sunjae; Quint, Evelien E; Clark-Cutaia, Maya N; Wu, Wenbo; Thompson, Valerie L; Crews, Deidra C; Purnell, Tanjala S; Thorpe, Roland J; Szanton, Sarah L; Segev, Dorry L; McAdams DeMarco, Mara A
IMPORTANCE/UNASSIGNED:Identifying the mechanisms of structural racism, such as racial and ethnic segregation, is a crucial first step in addressing the persistent disparities in access to live donor kidney transplantation (LDKT). OBJECTIVE/UNASSIGNED:To assess whether segregation at the candidate's residential neighborhood and transplant center neighborhood is associated with access to LDKT. DESIGN, SETTING, AND PARTICIPANTS/UNASSIGNED:In this cohort study spanning January 1995 to December 2021, participants included non-Hispanic Black or White adult candidates for first-time LDKT reported in the US national transplant registry. The median (IQR) follow-up time for each participant was 1.9 (0.6-3.0) years. MAIN OUTCOME AND MEASURES/UNASSIGNED:Segregation, measured using the Theil H method to calculate segregation tertiles in zip code tabulation areas based on the American Community Survey 5-year estimates, reflects the heterogeneity in neighborhood racial and ethnic composition. To quantify the likelihood of LDKT by neighborhood segregation, cause-specific hazard models were adjusted for individual-level and neighborhood-level factors and included an interaction between segregation tertiles and race. RESULTS/UNASSIGNED:Among 162 587 candidates for kidney transplant, the mean (SD) age was 51.6 (13.2) years, 65 141 (40.1%) were female, 80 023 (49.2%) were Black, and 82 564 (50.8%) were White. Among Black candidates, living in a high-segregation neighborhood was associated with 10% (adjusted hazard ratio [AHR], 0.90 [95% CI, 0.84-0.97]) lower access to LDKT relative to residence in low-segregation neighborhoods; no such association was observed among White candidates (P for interaction = .01). Both Black candidates (AHR, 0.94 [95% CI, 0.89-1.00]) and White candidates (AHR, 0.92 [95% CI, 0.88-0.97]) listed at transplant centers in high-segregation neighborhoods had lower access to LDKT relative to their counterparts listed at centers in low-segregation neighborhoods (P for interaction = .64). Within high-segregation transplant center neighborhoods, candidates listed at predominantly minority neighborhoods had 17% lower access to LDKT relative to candidates listed at predominantly White neighborhoods (AHR, 0.83 [95% CI, 0.75-0.92]). Black candidates residing in or listed at transplant centers in predominantly minority neighborhoods had significantly lower likelihood of LDKT relative to White candidates residing in or listed at transplant centers located in predominantly White neighborhoods (65% and 64%, respectively). CONCLUSIONS/UNASSIGNED:Segregated residential and transplant center neighborhoods likely serve as a mechanism of structural racism, contributing to persistent racial disparities in access to LDKT. To promote equitable access, studies should assess targeted interventions (eg, community outreach clinics) to improve support for potential candidates and donors and ultimately mitigate the effects of segregation.
PMCID:10877505
PMID: 38372985
ISSN: 2168-6114
CID: 5634032

Evolving Trends in Kidney Transplant Outcomes Among Older Adults: A Comparative Analysis Before and During the COVID-19 Pandemic

Li, Yiting; Menon, Gayathri; Wu, Wenbo; Musunuru, Amrusha; Chen, Yusi; Quint, Evelien E; Clark-Cutaia, Maya N; Zeiser, Laura B; Segev, Dorry L; McAdams-DeMarco, Mara A
BACKGROUND/UNASSIGNED:Advancements in medical technology, healthcare delivery, and organ allocation resulted in improved patient/graft survival for older (age ≥65) kidney transplant (KT) recipients. However, the recent trends in these post-KT outcomes are uncertain in light of the mounting burden of cardiovascular disease, changing kidney allocation policies, heterogeneity in candidates' risk profile, and the coronavirus disease 2019 pandemic. Thus, we examined secular trends in post-KT outcomes among older and younger KT recipients over the last 3 decades. METHODS/UNASSIGNED:We identified 73 078 older and 378 800 younger adult (aged 18-64) recipients using Scientific Registry of Transplant Recipients (1990-2022). KTs were grouped into 6 prepandemic eras and 1 postpandemic-onset era. Kaplan-Meier and Cox proportional hazards models were used to examine temporal trends in post-KT mortality and death-censored graft failure. RESULTS/UNASSIGNED:From 1990 to 2022, a 19-fold increase in the proportion of older KT recipients was observed compared to a 2-fold increase in younger adults despite a slight decline in the absolute number of older recipients in 2020. The mortality risk for older recipients between 2015 and March 14, 2020, was 39% (adjusted hazard ratio [aHR] = 0.61, 95% confidence interval [CI], 0.50-0.75) lower compared to 1990-1994, whereas that for younger adults was 47% lower (aHR = 0.53, 95% CI, 0.48-0.59). However, mortality risk during the pandemic was 25% lower (aHR = 0.75, 95% CI, 0.61-0.93) in older adults and 37% lower in younger adults (aHR = 0.63, 95% CI, 0.56-0.70) relative to 1990-1994. For both populations, the risk of graft failure declined over time and was unaffected during the pandemic relative to the preceding period. CONCLUSIONS/UNASSIGNED:The steady improvements in 5-y mortality and graft survival were disrupted during the pandemic, particularly among older adults. Specifically, mortality among older adults reflected rates seen 20 y prior.
PMCID:10624464
PMID: 37928483
ISSN: 2373-8731
CID: 5606682