key: cord-0760715-46htv4tl
authors: Hawley, Jessica E.; Sun, Tianyi; Chism, David D.; Duma, Narjust; Fu, Julie C.; Gatson, Na Tosha N.; Mishra, Sanjay; Nguyen, Ryan H.; Reid, Sonya A.; Serrano, Oscar K.; Singh, Sunny R. K.; Venepalli, Neeta K.; Bakouny, Ziad; Bashir, Babar; Bilen, Mehmet A.; Caimi, Paolo F.; Choueiri, Toni K.; Dawsey, Scott J.; Fecher, Leslie A.; Flora, Daniel B.; Friese, Christopher R.; Glover, Michael J.; Gonzalez, Cyndi J.; Goyal, Sharad; Halfdanarson, Thorvardur R.; Hershman, Dawn L.; Khan, Hina; Labaki, Chris; Lewis, Mark A.; McKay, Rana R.; Messing, Ian; Pennell, Nathan A.; Puc, Matthew; Ravindranathan, Deepak; Rhodes, Terence D.; Rivera, Andrea V.; Roller, John; Schwartz, Gary K.; Shah, Sumit A.; Shaya, Justin A.; Streckfuss, Mitrianna; Thompson, Michael A.; Wulff-Burchfield, Elizabeth M.; Xie, Zhuoer; Yu, Peter Paul; Warner, Jeremy L.; Shah, Dimpy P.; French, Benjamin; Hwang, Clara
title: Assessment of Regional Variability in COVID-19 Outcomes Among Patients With Cancer in the United States
date: 2022-01-04
journal: JAMA Netw Open
DOI: 10.1001/jamanetworkopen.2021.42046
sha: aad7cdd195ba7a563a0443f2508e3fe9973e27ec
doc_id: 760715
cord_uid: 46htv4tl

IMPORTANCE: The COVID-19 pandemic has had a distinct spatiotemporal pattern in the United States. Patients with cancer are at higher risk of severe complications from COVID-19, but it is not well known whether COVID-19 outcomes in this patient population were associated with geography. OBJECTIVE: To quantify spatiotemporal variation in COVID-19 outcomes among patients with cancer. DESIGN, SETTING, AND PARTICIPANTS: This registry-based retrospective cohort study included patients with a historical diagnosis of invasive malignant neoplasm and laboratory-confirmed SARS-CoV-2 infection between March and November 2020. Data were collected from cancer care delivery centers in the United States. EXPOSURES: Patient residence was categorized into 9 US census divisions. Cancer center characteristics included academic or community classification, rural-urban continuum code (RUCC), and social vulnerability index. MAIN OUTCOMES AND MEASURES: The primary outcome was 30-day all-cause mortality. The secondary composite outcome consisted of receipt of mechanical ventilation, intensive care unit admission, and all-cause death. Multilevel mixed-effects models estimated associations of center-level and census division–level exposures with outcomes after adjustment for patient-level risk factors and quantified variation in adjusted outcomes across centers, census divisions, and calendar time. RESULTS: Data for 4749 patients (median [IQR] age, 66 [56-76] years; 2439 [51.4%] female individuals, 1079 [22.7%] non-Hispanic Black individuals, and 690 [14.5%] Hispanic individuals) were reported from 83 centers in the Northeast (1564 patients [32.9%]), Midwest (1638 [34.5%]), South (894 [18.8%]), and West (653 [13.8%]). After adjustment for patient characteristics, including month of COVID-19 diagnosis, estimated 30-day mortality rates ranged from 5.2% to 26.6% across centers. Patients from centers located in metropolitan areas with population less than 250 000 (RUCC 3) had lower odds of 30-day mortality compared with patients from centers in metropolitan areas with population at least 1 million (RUCC 1) (adjusted odds ratio [aOR], 0.31; 95% CI, 0.11-0.84). The type of center was not significantly associated with primary or secondary outcomes. There were no statistically significant differences in outcome rates across the 9 census divisions, but adjusted mortality rates significantly improved over time (eg, September to November vs March to May: aOR, 0.32; 95% CI, 0.17-0.58). CONCLUSIONS AND RELEVANCE: In this registry-based cohort study, significant differences in COVID-19 outcomes across US census divisions were not observed. However, substantial heterogeneity in COVID-19 outcomes across cancer care delivery centers was found. Attention to implementing standardized guidelines for the care of patients with cancer and COVID-19 could improve outcomes for these vulnerable patients.

Approved Project Title

Jessica Hawley and Clara Hwang 1 (a) Manuscript title Assessment of US Regional Variability in COVID-19 Outcomes Among Patients With Cancer 3 Objectives State specific objectives, including any prespecified hypotheses H0: Clinical outcomes and US census region are independent in patients with cancer and COVID-19 in adjusted models (no difference in death (30-day all-cause mortality) or composite endpoint (death, rate of ICU admission, rate of mechanical ventilation). HA: Clinical outcomes and US census region are associated in patients with cancer and COVID-19. 4 Study design Retrospective, multi-center cohort study 5 Setting

Cohort study of patients with current or past history of cancer and lab-confirmed SARS-CoV-2 infection from >120 participating CCC19 sites. Data analyzed from March 15, 2020 -November 30, 2020. Registry built and maintained as electronic REDCap database at VUMC. 6 Participants (a) Give the eligibility criteria, and the sources and methods of selection of participants. Describe methods of follow-up Subjects included in this analysis were those with laboratory-confirmed SARS-CoV-2 infection and past or current history of invasive malignancy, from the U.S., and entered into the CCC19 database between 3/17/2020 and 11/30/2020 with follow-up data reported through 12/31/2020. We excluded records of presumptive COVID-19 cases, patients < 18 years of age, patients with incomplete follow-up data, cases of non-invasive cancers and premalignant conditions or non-melanoma noninvasive skin cancers, cases from outside the U.S., and records with quality score >4. (b) For matched studies, give matching criteria and number of exposed and unexposed Not a matched study. Variables of interest are defined by the CCC19 database registry and entered by each local site's data abstractor.

-race/ethnicity self-reported (5 categories) -smoking (never v. ever v. missing) -obesity (y/n/missing) -specific comorbidities (CV, pulm, renal, DM, missing) -cancer status (active and responding, active and stable, active and progressing vs. remission/NED, unknown, missing) -ECOG (0, 1, ≥ 2, unknown, missing) -type of malignancy (solid vs. heme) -modality of active anti-cancer therapy (categorical: none, cytotoxic chemo, immunotherapy, targeted, endocrine, locoregional, other, missing/unknown) -COVID-19 tx (categorical: use JCO paper groupings) -Rural-urban status (center-level: ordinal -1, 2, 3) -Neighborhood (cont: use center-level SVI) -Population density by US census subregion (continuous) -New COVID19 ave daily case over 3 time intervals by region (continuous) 9 Bias Describe any efforts to address potential sources of bias Adjustment for covariates in multivariable models.

10 Study size Explain how the study size was arrived at Case volume dependent on data abstracters at each site.

11 Quantitative variables Explain how quantitative variables will be handled in the analyses. If applicable, describe which groupings will be chosen and why

As per above 12 Statistical methods (a) Describe all statistical methods, including those to be used to control for confounding Covariates (listed above under potential confounders) and binary outcomes will be summarized across 9 census subregions using standard descriptive statistics. Multivariable generalized linear mixed-effects models (with a logit link for binary outcomes and center-level random effects) will be used to (1) estimate adjusted covariate-outcome associations and (2) estimate adjusted subregion-level outcome rates overall and at 3-month time intervals. For secondary outcomes, the model will include an offset for (log) follow-up time. A list of potential tables and figures is provided below. (b) Describe any methods that will be used to examine subgroups and interactions None (c) Explain how missing data will be addressed Multiple imputation will be used to impute missing and unknown data for all variables included in the analysis, with some exceptions: unknown ECOG performance score and unknown cancer status will not be imputed and treated as a separate category in analyses. Imputation will be performed on the largest dataset possible (that is, after removing test cases and other manual exclusions, but before applying specific exclusion criteria). At least 10 imputations will be generated.

(d) If applicable, explain how loss to follow-up will be addressed Excluded if no data on 30-day follow-up form. We will extend follow-up time (through 12/31/2020) to ensure that follow-up for the primary outcome, 30-day mortality, is complete to the best of our ability. 

Once the project design and SAP have been approved for your project, this document will be used to specify the exact outcomes and variables to be used in your analysis. Please provide as much information as you can, including the existing variable name if you know it. Use of existing variables will decrease the amount of time that it takes to get your project to the analysis phase, but we will endeavor to add any needed derived variables based on your project needs. Existing variables can be found in two places within the GitHub repo: the Data Dictionary ("CCC19_DataDictionary.csv"), which includes the native variables found in the survey, and the list of Derived Variables ("CCC19_Derived_Variables_Spreadsheet.xlsx"). The composite outcome reflected the occurrence of any of the following: admission to an intensive care unit, receipt of mechanical ventilation, and total all-cause mortality. Analyses of the composite outcome were limited to 4,561 patients within non-missing data. b Odds ratios greater than 1 indicate higher odds of 30-day all-cause mortality. c Odds ratios greater than 1 indicate higher odds of admission to an intensive care unit, receipt of mechanical ventilation, or total all-cause mortality. d Adjusted for age, sex, race and ethnicity, smoking status, obesity, cardiovascular comorbidities, pulmonary comorbidities, renal disease, diabetes mellitus, type of malignancy, cancer status, Eastern Cooperative Oncology Group performance status, anti-COVID-19 treatments, and month of COVID-19 diagnosis. P values for evaluating the null hypothesis of equality in odds ratios across census divisions (8 degrees of freedom): 30-day mortality, 0.42; composite outcome, 0.73.

RN (Mount Carmel Health System

MD (Rutgers Cancer Institute of New Jersey at Rutgers Biomedical and Health Sciences

Scott Cancer Center at LSU Health Sciences Center

APC (ThedaCare Cancer Care

MD (Tisch Cancer Institute at the Icahn School of Medicine at Mount Sinai

MD (Tufts Medical Center Cancer Center

MD (UC Davis Comprehensive Cancer Center at the University of California at

MD (Wake Forest Baptist Comprehensive Cancer Center

The Centers for Disease Control and Prevention's Social Vulnerability Index (SVI) indicates the relative vulnerability of every U.S. Census tract (or county) according to ranks on 15 social factors in four domains: socioeconomic status, household composition and disability, minority status and language, and housing and transportation. An overall percentile rank (0-1) for the county where each center is located was used in this analysis. A SVI ranking of 0.85 indicates that 85% of tracts (or counties) in the state or nation are less vulnerable than the tract of interest and that 15% of tracts (or counties) in the state or nation are more vulnerable.The U.S. Department of Agriculture's Economic Research Service's 2013 Rural-Urban Continuum Code (RUCC) classifies all counties in the U.S. by their official metro-nonmetro status, as defined by the Office of Management and Budget, with further breakdown by population resulting in nine RUCC codes (1-9). In this coding scheme, 1 represents counties in metropolitan areas with populations of 1 million, and 9 represents counties that are completely rural with populations of <2500, not adjacent to a metro area.

Jun-Aug, 2020 2.62 (0.86-8.01) CI, confidence interval, OR, odds ratio. a P value for evaluating the null hypothesis of equality in odds ratios across month of COVID-19 diagnosis: 0.13 (2 degrees of freedom).