Variability in Excess ESKD Patient Mortality Among States During the Pandemic

Background: The incidence of death among patients with ESKD in the United States has been significantly elevated during the COVID-19 pandemic, relative to recent historical levels. In the general population, individual states have exhibited markedly variable COVID-19 death rates, possibly reflecting policy decisions concerning pandemic management. However, little is known about regional variability in the excess mortality of ESKD patients. We analyzed excess mortality among both prevalent dialysis and kidney transplant (KT) patients between March 15, 2020, and June 30, 2021. Method(s): We analyzed national data extracted from the Centers for Medicare and Medicaid Services (CMS) End Stage Quality Reporting System (EQRS). For each epidemiologic week from week 1 of 2018 to week 26 of 2021, we identified prevalent dialysis and KT patients at the beginning of each week, and the incidence of all-cause death among them during the week. For each combination of state and kidney replacement therapy, we estimated excess mortality during the pandemic (definition: week 12 of 2020 and thereafter), using a logistic regression model of death among patient-weeks, with adjustment for age, sex, race. Result(s): From 732,063 dialysis patients in 50 states, there were 129,095 pandemicera deaths;and from 238,265 KT patients in 41 states, there were 11,256 pandemicera deaths. State-level excess mortality ratios in dialysis patients ranged from 0.96 to 1.34, with 40% of states having ratios significantly greater than 1.1. State-level excess mortality ratios in KT patients ranged from 1.02 to 2.73, with 61% of states having ratios significantly greater than 1.1. Conclusion(s): States exhibited wide variability in excess ESKD patient mortality, with roughly 3-fold wider variability in prevalent KT patients than in dialysis patients.

Use of Non-Approved and Approved Treatments for COVID-19 by CKD Status in Medicare

Background: Hydroxychloroquine, chloroquine, and ivermectin gained popularity for treatment of COVID-19 in 2020. Remdesivir was approved for treating hospitalized COVID-19 in late 2020. We studied the uptake of these drugs early in the pandemic in a 5% sample of Medicare fee-for-service beneficiaries with and without CKD. Method(s): We examined the percentage of beneficiaries receiving >=1 Part D covered prescription for hydroxychloroquine or chloroquine and ivermectin in each month of 2020. Among first COVID-19 hospitalizations from November 2020-June 2021, we examined the percentage receiving remdesivir using ICD-10-PCS. Analyses included those aged >=66 years without ESRD;CKD was defined by >=1 inpatient or >=2 outpatient diagnoses. Result(s): Use of hydroxychloroquine and chloroquine increased in March 2020 and then subsided in the ensuing months, remaining slightly elevated though 2020 (Figure A). Receipt of these drugs was higher in patients with CKD than in those without. Ivermectin use was uncommon in both groups before spiking in December 2020 (Figure B). Among COVID-19 inpatients, 55% without CKD and 44% with CKD received remdesivir, which was used more often in men than in women, less often in Blacks than in Whites or Hispanics, and less often in those with the low-income subsidy than in those without. Conclusion(s): In 2020, Medicare beneficiaries with and without CKD showed similar spiking patterns in use of the approval-revoked or non-approved drugs hydroxychloroquine/chloroquine (in March) and ivermectin (in December). Through June 2021, remdesivir was used less in patients with CKD than in those without for hospitalized COVID-19, likely because the FDA recommends not using remdesivir if eGFR is <30 mL/min. Lower income and Black patients were less likely to receive remdesivir than others.

Undercounting of COVID-19 Vaccinations in Fee-for-Service Medicare Claims

Background: Older individuals and those with certain underlying conditions were among the earliest groups offered COVID-19 vaccinations. While patients with ESKD did not initially receive priority, a federal program permitted vaccinations to be administered in dialysis clinics starting in March 2021. We studied early uptake of COVID-19 vaccinations in Medicare fee-for-service beneficiaries with ESKD. Method(s): We included beneficiaries aged >=18 years with ESKD on December 1, 2020 from the US Renal Data System. Vaccinations covered by Medicare were identified using CPT codes. The cumulative monthly incidence of first vaccination dose through June 2021 was compared by modality (HD, PD, transplant) and stratified by age and race/ ethnicity. Death was treated as a competing risk. Result(s): By June 30, 2021, the cumulative incidence of receiving a Medicare-covered first vaccination dose was <40% in patients receiving HD (Figure A), well under the estimate reported by dialysis facilities to the CDC by this date (72%). Although caution is required, some interpretation of the Medicare vaccination data may still be permitted. After the allocation of vaccines to dialysis clinics, Medicare-covered vaccinations surged in patients receiving HD relative to the other modalities. In patients receiving HD, uptake of Medicare-covered vaccinations was initially highest among those aged >=65 years and then surged in younger patients following the federal vaccine allocation (Figure B). Conclusion(s): COVID-19 vaccination rates are severely underestimated using Medicare administrative data. It is unclear whether missingness of vaccination data is differential by demographic groups, such as race/ethnicity. Inferences based on these data should be made with caution.

Changes in timing and preparedness for dialysis initiation in the era of COVID-19

Background: How the COVID-19 pandemic altered aspects of dialysis initiation, such as eGFR at initiation, selection of peritoneal dialysis (PD), and, in patients initiating hemodialysis (HD), use of a central venous catheter (CVC), is not fully understood. Methods: We analyzed the most recently updated quarterly USRDS data available. Using Poisson and logistic regression, we studied weekly changes in eGFR at dialysis initiation, use of PD (versus HD), and use of incident CVCs, overall and by strata of race, during the first half of 2020 compared to a forecast of 2020, had 2017-2019 historical trends continued. Results: Mean eGFR at dialysis initiation decreased by 0.33 mL/min/1.73 m2 in weeks 19-22, compared with historical trends;non-Hispanic Black patients exhibited the largest decrease, at 0.61 mL/min/1.73 m2. The odds of initiating dialysis with eGFR <10 ml/min/ 1.73m2 were highest during weeks 19-22 (May;OR 1.14, 1.05-1.17), corresponding to an absolute increase of 2.9%. Although initiation of both HD and PD fell, PD fell less, such that the odds of initiating PD (versus HD) were 24% higher (OR 1.24, 1.14-1.34) in weeks 11-14. Odds of initiating HD with a CVC increased by 30% (OR 1.30, 1.20-1.41) in weeks 15-18, representing an absolute increase of 3.3%. Conclusions: In the first half of 2020, eGFR at dialysis initiation fell, most prominently in non-Hispanic Blacks. During the initial wave of the pandemic, odds of utilizing PD, compared with HD, increased by nearly 25%, and odds of using a CVC at HD initiation increased by 30%.

Association of the COVID-19 pandemic with ESKD incidence

Background: How the COVID-19 pandemic altered ESKD incidence, dialysis initiation, and preemptive kidney transplantation is unknown. Methods: Using Centers for Medicare & Medicaid Services data, we investigated the incidence of ESKD, dialysis initiation, and preemptive kidney transplantation by week during the first half of 2020. Using Poisson regression, we compared findings in 2020 to a forecast of 2020, had 2017-2019 historical trends continued, overall and by strata of age and race. Results: Mean weekly counts of patients with new ESKD are shown in the Figure. Incidence of ESKD dropped dramatically in 2020 compared with the expected incidence, particularly during epidemiologic weeks 15-18 (April;incidence rate ratio [IRR] 0.75, 95% CI 0.73-0.78), before approaching pre-pandemic levels in weeks 23-26 (June;IRR 0.93, 0.90-0.95). Across age groups, the decrease was most pronounced during weeks 15-18 among individuals aged ≥75 years (IRR 0.69, 0.66-0.73, compared with individuals aged 45-64 years, IRR 0.80, 0.77-0.84). In terms of race, the decrease was least notable among non-Hispanic Blacks (IRR 0.85, 0.81-0.89) and was most pronounced in non-Hispanic Whites (IRR 0.72, 0.69-0.74) and Hispanics (IRR 0.73, 0.69-0.78). Dialysis initiation reached a nadir during weeks 15-18 (IRR 0.76, 0.74-0.78), and preemptive kidney transplantation decreased even more strikingly during this period (IRR 0.56, 0.46-0.67). Conclusions: During the first wave of the COVID-19 pandemic in 2020, the number of patients starting treatment for ESKD fell to a level not observed since 2011. Changes in ESKD incidence and utilization of treatment modalities may reflect differential access to care.

Associations of dialysis modality and setting with incidence of COVID-19 diagnosis and hospitalization

Background: The novel coronavirus 2019 (COVID-19) pandemic has resulted in substantial morbidity and mortality among patients undergoing maintenance dialysis. Patients performing home hemodialysis (HHD) or peritoneal dialysis (PD) may be able to minimize exposure to the community, thus lowering risk of COVID-19 infection. We assessed whether HHD and PD were associated with lower risks of COVID-19 infection and hospitalization, compared to in-facility hemodialysis (IHD). Methods: We analyzed Medicare Parts A and B claims accrued during 2020. For each epidemiologic week from week 12 (beginning March 15) to week 37 (September 6), we identified patients with a Medicare-covered outpatient dialysis treatment during the preceding 7 days. We stratified patients into cohorts of IHD, HHD, and PD;we limited the IHD cohort to patients without residency in a skilled nursing facility during the 28 days preceding the epidemiologic week. During each week, we estimated the incidence of COVID-19 infection and COVID-19 hospitalization, per Medicare claims with ICD-10-CM diagnosis code U07.1. Using logistic regression with adjustment for demography, comorbidity, and state, we estimated odds ratios of outcomes during weeks 12-22, 23-33, and 34-37. Results: Incidence of COVID-19 infection (figure) and COVID-19 hospitalization peaked twice: during weeks 14-16 and weeks 29-30. During weeks 12-22, adjusted odds ratios (AORs) of COVID-19 infection for HHD versus IHD and PD versus IHD were 0.55 (95% CI, 0.43-0.71) and 0.52 (0.46-0.58), respectively. During weeks 23-33, corresponding AORs were 0.63 (0.50-0.78) and 0.63 (0.57-0.69). Conclusions: Both home dialysis modalities were associated with similarly lower risks of COVID-19 infection and hospitalization. Nephrologists and dialysis provider should consider counseling patients about potentially lower risk of infectious respiratory disease with home dialysis.

Excess COVID and non-COVID mortality among patients with ESKD in 2020

Background: Patients with ESKD are at high risk of mortality from COVID-19. The extent to which increased mortality in the ESKD population in 2020 was related to COVID-19 vs other causes in the setting of disruption of healthcare delivery is not clear. Methods: We used the Death Notification Form (CMS-2746) to examine excess allcause and COVID-related mortality in Jan through Aug of 2020 among the whole ESKD population and by race/ethnicity adjusting for age and comorbidity. We further examined causes of non-COVID-related mortality in 2020 compared with 2017-2019. Results: All-cause mortality increased by 13.1% during Jan-Aug 2020 compared with mortality during the same period in 2017-2019. Peak overall and excess mortality occurred during Apr (Figure), when 14.7% of all deaths were attributed to COVID-19. COVID-related deaths declined to a nadir of 6.2% of all deaths in June and then increased again in Jul and Aug. Excess mortality was approximately twice as high among Black and Hispanic patients as among whites. Between Feb and Apr, there was substantial excess non-COVID mortality in addition to COVID mortality, whereas most excess mortality during May to Aug was related to COVID-19. There were 4310 excess deaths during Feb-Apr 2020. 1576 (37%) were due to COVID-19, and 517 (12%) were attributed to pneumonia. Thus, approximately half of the excess mortality was not due to COVID-19, including 1635 excess deaths due to cardiac arrest, cause unknown. Conclusions: Patients with ESKD experienced substantial excess mortality in 2020 relative to prior years that affected Black and Hispanic patients disproportionately. Approximately half of the excess mortality was likely caused by COVID-19. There was also excess mortality in the early phase of the pandemic that was not attributed to COVID-19.