Abstract
Background:
Reports of poor outcomes among older adults with COVID-19 may have changed patient perceptions of Do-Not-Resuscitate (DNR) orders or caused providers to pressure older adults into accepting DNR orders to conserve resources.
Objective:
We determined early-DNR utilization during COVID-19 surges compared with nonsurge periods among nonsurgical adults ≥75 and its connection to hospital mortality.
Methods:
We conducted a retrospective cohort study among adults ≥75 years using the California Patient Discharge Database 2020. The primary outcome was early-DNR utilization. Control cohorts included nonsurgical adults <75 years in 2020 and nonsurgical adults ≥75 in 2019. Multiple causal inference methods were used to address measured and unmeasured confounding.
Results:
A total of 487,955 adults ≥75 years were identified, with 233,678 admitted during COVID-19 surges. Older adults admitted during surges had higher rates of early-DNR orders (30.1% vs. 29.4%, absolute risk differences = 0.7, 95% confidence interval [CI]: 0.5–1.0) even after adjusting for patient case-mix (adjusted odds ratio [aOR] = 1.02, 95% CI: 1.01–1.04). Patients with early-DNR orders experienced higher hospital mortality (15.5% vs. 4.8%, aOR = 3.96, 95% CI: 3.85–4.06). Difference-in-difference analyses demonstrated that adults <75 years in 2020 and adults ≥75 years in 2019 did not experience variation in early-DNR utilization.
Conclusions:
Older adults had slightly higher rates of early-DNR orders during COVID-19 surges compared with nonsurge periods. While the difference in early-DNR utilization was small, it was linked to higher odds of death. The increase in early-DNR use only during COVID-19 surges and only among older adults may reflect changes in patient preferences or increased pressure on older adults stemming from provider fears of rationing during COVID-19 surges.
Keywords: ageism, Crisis Standards of Care, COVID-19, DNR orders
Introduction
Early information during the SARS-CoV-2 (COVID-19) pandemic, which was later found to be misleading, suggested alarmingly high mortality for patients with COVID-19 who required mechanical ventilation (MV).1,2 The lack of concrete scientific information, headline grabbing reports of overwhelmed hospitals and morgues at capacity, and daily death counts led to widespread fear among both the public and health care providers of how bad the COVID-19 pandemic could get.3–5 These fears were coupled with concerns that the surge of critically ill patients would require activation of Crisis Standards of Care (CSC) with rationing of resources like ventilators.6–8
Given widespread fears, palliative care specialists assumed critical roles in the pandemic response in discussing advanced care planning and Do-Not-Resuscitate (DNR) orders. In multiple hospitals, palliative care specialists served on CSC triage and ethics committees, assisted in family communication when visitors were banned, and became essential in discussing advanced directives to avoid use of scarce resources for patients who did not want aggressive interventions. While it was recognized that fear of MV might contribute to more patients refusing life-prolonging therapies, there was also concern that vulnerable patients might be pressured or coerced into agreeing to DNR orders to conserve limited resources, a form of implicit rationing.9–12 Multiple institutions considered universal DNR orders for patients with COVID-19.11–13 However, palliative care specialists advocated for an individualized approach. Eventually, the United States Office of Civil Rights (OCR) held that CSC rationing protocols could not categorically exclude patients based only on age, disability status, gender, race, or other demographic characteristics.14–16
Moreover, OCR also stipulated that patients could not be pressured or coerced to accept DNR orders to conserve resources.16 However, anecdotal reports suggest that some vulnerable groups may have been inappropriately pressured to accept DNR orders during COVID-19 surges.10,17–22
Despite the concerns about patient fears and resource constraints, little is known about how DNR orders were utilized during COVID-19 surges in 2020. Of particular concern is the use of DNR orders among older adults as they likely had greater fear of MV based on media reports and were potentially subject to greater pressure from providers to consider limitations to care. We conducted a retrospective cohort study investigating use of hospital DNR orders among older adults during COVID-19 surges. We hypothesized that admission during a COVID-19 surge compared with a nonsurge period should not be associated with differential early-DNR utilization among older adults.
Methods
Please see Online Supplementary Methods S1 for full details.
Study design
We conducted a retrospective cohort study using the California Patient Discharge Data (PDD) 2020, an administrative discharge database with 100% of nonfederal discharges from California.23,24 The PDD is the largest all-payer database with an independently validated indicator of DNR status within the first 24 hours of admission (early-DNR).25,26
Patients
The primary cohort included nonsurgical patients ≥75 years admitted in 2020 (Fig. 1).27 Patients with missing admission dates and DNR indicators were excluded. Two comparator cohorts were also used: (1) nonsurgical patients <75 years admitted in 2020 to determine if any differences in DNR utilization were age-related, and (2) nonsurgical patients ≥75 years admitted in 2019 to determine if any differences in DNR utilization were unique to 2020 or represented annual seasonal variations.
FIG. 1.
Consort diagram. The California Patient Discharge Database 2020 contained records for 2,629,541 hospitalizations. Surgical patients, patients with missing early-DNR indicator, and patients <75 were excluded from the primary cohort resulting in 487,955 nonsurgical adults ≥75 years of age in the primary cohort. In the primary cohort, 233,326 (47.8%) were admitted during COVID-19 surges and 254,629 (52.2%) were admitted during nonsurge periods. COVID-19, SARS-CoV-2; DNR, Do-Not-Resuscitate.
Exposure
The primary exposure was admission during COVID-19 surge months (April, June, July, August, November, and December) compared with nonsurge months (January, February, March, May, September, and October) during 2020. Surge periods were based on the changes in the total number of COVID-19 cases per month and the average daily COVID-19 hospital census per month (Supplementary Table S1).28
Outcome
The primary outcome was early-DNR status identified through the PDD indicator or through International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) diagnosis code Z66 present on admission.25,29 The focus was on DNR status at admission as later-DNR status (DNR orders placed >24 hours after admission) could reflect clinical decline rather that initial patient preference. Early-DNR orders reflect severity of prehospital comorbid illness more than acute illness, supporting the hypothesis that DNR utilization should be stable during surge and nonsurge times.30 To gauge the clinical impact of early-DNR orders, the secondary outcome was hospital mortality based on DNR status.
Statistical analysis
Continuous and categorical variables are reported as means with standard deviations (SD), medians with interquartile ranges, and percentages and absolute risk differences (ARD). We used hierarchical logistic regression with the hospital as a random intercept to determine the adjusted odds ratio (aOR) for early-DNR status and hospital mortality between surge and nonsurge periods. Models were adjusted for patient demographics (age, gender, race, and ethnicity), individual Elixhauser chronic comorbidities, and acute organ failures present on admission (Supplementary Table S2).31–37 Similar multilevel models were used to assess hospital mortality. Patients with missing data were excluded from regression models.
Causal inference sensitivity analyses
Due to imbalance in covariates between surge and nonsurge months, we conducted a sensitivity analysis weighting the analysis by the inverse probability of treatment (IPTW) to determine the aOR of early-DNR status in the weighted sample where “treatment” was the admission time period (surge vs. nonsurge). IPTW led to standardized mean differences between exposed and unexposed of <0.001 for all covariates (Supplementary Table S3). Additionally, we conducted two separate difference-in-difference analyses using the comparator cohorts to investigate whether any potential differences in DNR utilization between surge and nonsurge months were unique to older adults in 2020. We compared differences in early-DNR utilization among adults ≥75 years in 2020 to differences in early-DNR utilization among adults <75 years in 2020 to determine if any differences were unique to the older population. We also compared differences in early-DNR utilization during surge versus non-surge periods between adults ≥75 years in 2020 to differences for adults ≥75 years in 2019 to determine if differences were unique to the COVID-19 time period (2020) or reflected annual seasonal variation.
Statistical sensitivity analyses
We conducted several sensitivity analyses to address potential statistical issues. First, given concerns about potential misclassification bias with use of ICD-10-CM billing codes, we repeated the primary analysis where early-DNR status was only defined based on the unique identifier in the PDD. Second, we repeated the primary analysis but included May as a surge month. Lastly, we attempted to address potential model overfit based on the large sample size and large number of covariates with potential collinearity in the primary model. We employed two variable reduction methods based on the Least Absolute Shrinkage and Selection Operator (LASSO) method (see Online Supplement for details) to address potential model overfit and collinearity.38
Statistical testing was two-tailed with an a priori alpha = 0.05 using SAS v9.4 (Cary, NC). The study was deemed exempt by the Colorado Multiple Institutional Review Board (Aurora, CO, Protocol 20–2514, Long-Term Outcomes for Critical Illness, Initial Review Date: October 30, 2020).
Results
Patients
There were 487,955 admissions for adults ≥75 years in California in 2020. Among the older adults, 233,326 (47.8%) adults were admitted during COVID-19 surge periods and 254,629 (52.2%) were admitted in nonsurge months (Fig. 1). Patients admitted during COVID-19 surge months were of similar age and similar comorbidity risk profiles (Table 1, Supplementary Table S4). A slightly lower percentage of patients admitted during COVID-19 surge periods were female compared with those admitted during nonsurge periods. 65.3% of patient older adults admitted during COVID-19 surge periods identified as White and 6.3% identified as Black compared with 66.2% and 6.2% during nonsurge periods.
Table 1.
Patient Characteristics
| Characteristica | Surge admission (n = 233,326) | Nonsurge admission (n = 254,629) |
|---|---|---|
| Demographics | ||
| Mean age, year, (SD) | 83.7 (6.2) | 83.8 (6.2) |
| Female (%) | 126,608 (54.2) | 141,257 (55.4) |
| Race (%) | ||
| White | 152,499 (65.3) | 168,689 (66.2) |
| Black | 14,822 (6.3) | 15,909 (6.2) |
| Asian | 26,523 (11.4) | 29,874 (11.7) |
| Other | 39,819 (17.0) | 40,462 (15.9) |
| Hispanic (%) | 47,333 (20.6) | 48,389 (19.2) |
| Clinical conditions | ||
| Median HCUP Comorbidity Mortality Predictor score (IQR)b | 14.0 (2.0–25.0) | 14.0 (2.0–25.0) |
| Cancer (%) | 26,732 (11.5) | 29,632 (11.6) |
| Chronic heart failure (%) | 81,170 (34.8) | 93731 (36.8) |
| Chronic lung disease (%) | 57,799 (24.8) | 67,843 (26.6) |
| Dementia (%) | 62,368 (26.7) | 64,484 (25.3) |
| Severe chronic liver disease (%) | 3016 (1.3) | 3282 (1.3) |
| Severe chronic renal disease (%) | 25,583 (10.0) | 23,470 (10.0) |
| COVID-19 (%) | 32,527 (13.9) | 7618 (3.0) |
| Mechanical ventilation (%) | 13,942 (6.0) | 12,483 (4.9) |
Please see Supplementary Table S4 for full patient characteristics, including comorbidities and acute organ failures present on admission.
The HCUP Comorbidity Mortality Predictor is a score calculated from a composite of 38 Elixhauser comorbidities and is designed to predict in hospital mortality. The score ranges from −45 to +171 with higher numbers indicating a higher probability of death. While the composite score, as a measure of comorbid illness, is presented in this table, individual Elixhauser comorbidities were included in statistical models to capture the prehospital likelihood of death. More information can be found at https://www.hcup-us.ahrq.gov/toolssoftware/comorbidityicd10/comorbidity_icd10.jsp
COVID-19, SARS-CoV-2 infection; HCUP, Health care Cost and Utilization Project; IQR, interquartile ranges; SD, standard deviations.
Outcomes
Among adults ≥75 years, 30.1% of patients admitted during COVID-19 surge periods had an early-DNR order compared with 29.4% during nonsurge periods (ARD 0.7, 95% confidence interval [CI]: 0.5–1.0) (Table 2). The ARD translates to an excess 1650 early-DNR orders (95% CI: 1230–2070) for older adults during COVID-19 surge periods. After adjusting for patient demographics, comorbidities, and severity of illness, older adults admitted during COVID-19 surge periods had higher odds of an early-DNR order compared with those admitted during nonsurge periods (aOR = 1.02, 95% CI: 1.01–1.04) (Table 2, Supplementary Table S5). Older adults with an early-DNR order also experienced significantly higher hospital mortality compared with patients without an early-DNR order during COVID-19 surges (15.5% vs. 4.8%, aOR = 3.96, 95% CI: 3.85–4.06) (Supplementary Table S6). Additionally, older adults with an early-DNR order had higher mortality during COVID-19 surges compared with nonsurge periods (17.4% vs. 13.8%, aOR = 1.20, 95% CI: 1.16–1.24) (Supplementary Table S7).
Table 2.
Early-Do-Not-Resuscitate Utilization During Surge Verses Non-Surge Periods
| Cohort | Surge admission (n = 233,678) | Non-surge admission (n = 254,953) | ARD (95% CI) | aOR for early-DNR (95% CI) | Difference-in-difference p-valuea |
|---|---|---|---|---|---|
| Adults ≥75 years in 2020 (%) | 70,155 (30.1) | 74,759 (29.4) | 0.7 (0.5–1.0) | 1.02 (1.01–1.04)b | — |
| Adults <75 years in 2020 (%) | 42,292 (6.4) | 43,269 (6.5) | −0.02 (−0.10–0.06) | 0.99 (0.98–1.01)c | <0.0001 |
| Adults ≥75 years in 2019 (%) | 77,280 (28.2) | 80,986 (28.4) | −0.2 (−0.4–0.1) | 1.01 (0.99–1.02)d | <0.0001 |
Difference-in-difference analysis compares the difference in early-DNR utilization between surge and nonsurge periods for the two comparator groups with the primary cohort of adults ≥75 years. The p-value is the p-value for the interaction term.
See Supplementary Table S5 for full model results.
See Supplementary Table S7 for full model results.
See Supplementary Table S8 for full model results.
aOR, adjusted odds ratio; ARD, absolute risk differences; CI, confidence interval; DNR, Do-Not-Resuscitate.
Causal inference sensitivity analyses
IPTW achieved strong covariate balance between surge and nonsurge periods (Supplementary Table S3). IPTW regression results matched the primary analysis and indicated that adults ≥75 years admitted during surge periods had higher odds of early-DNR orders compared with nonsurge times (aOR = 1.02, 95% CI: 1.01–1.04). The two difference-in-difference analyses supported the conclusion that the differential early-DNR utilization between surge and nonsurge periods was unique to older adults and unique to 2020. Adults <75 years in 2020 (Supplementary Table S8) had minimal differences in early-DNR utilization by surge versus nonsurge period (6.4% vs. 6.5%, ARD = −0.02, 95% CI: −0.10 to 0.06) even after adjusting for patient case-mix (Table 2, Fig. 2, Supplementary Table S9). Similarly, adults ≥75 in 2019 (Supplementary Table S10) had no difference in early-DNR utilization (28.2% vs. 28.4%, ARD = −0.2, 95% CI: −0.4 to 0.1) (Table 2, Fig. 2, Supplementary Table S11). The difference in DNR utilization among older adults in 2020 was significantly greater than younger adults in 2020 and older adults in 2019.
FIG. 2.
Monthly early-DNR utilization across three patient cohorts. The primary cohort consisted of nonsurgical patients ≥75 years admitted in 2020 (blue). The two comparator cohorts included (1) nonsurgical patients <75 years admitted in 2020 (orange) to evaluate for age-specific differences and (2) nonsurgical patients ≥75 years admitted in 2019 (green) to determine if differences in early-DNR utilization were unique to 2020. COVID-19 surge months are highlighted in red. Significant variability in early-DNR utilization during surge versus nonsurge months was present among older adults in 2020, but no significant differences were seen for younger adults in 2020 or older adults in 2019.
Statistical sensitivity analyses
When DNR was defined only with the unique indicator in the PDD, DNR utilization was higher during surge periods versus nonsurge periods (24.8% vs. 24.2%) even after adjusting for patient case mix (Supplementary Table S12). When May was included as a surge month, we observed similar increases in early-DNR utilization during surge months (30.1% vs. 29.2%, aOR = 1.03, 95% CI: 1.01–1.04) (Supplementary Table S13). When LASSO variable reduction methods were used to address model overfit, we observed similar results to the primary analysis (Supplementary Tables S14 and S15).
Discussion
Reports of overflowing ICUs and high mortality during COVID-19 surges in 2020 led to significant fear related to the survivability of COVID-19 and rationing of critical care resources like ventilators. Assessing patients' goals of care and considerations for DNR orders gained greater significance to conserve scarce resources. During this time, palliative care specialists advocated for an individualized approach without resorting to blanket or pressured DNR orders but concerns were raised about some groups of patients being pressured into accepting DNR orders.10,21 In this study, older adults were more likely to have an early-DNR order during COVID-19 surges compared with nonsurge periods. No difference in early-DNR utilization was identified among younger adults in 2020 or in older adults in 2019 suggesting that the observation was limited to adults ≥75 years of age in 2020. The difference in early-DNR utilization was small but robust to multiple sensitivity analyses and associated with significantly higher mortality.
When interpreting the results of this study, it must be recognized that the difference in early-DNR utilization was small but significant. No clear threshold for clinical significance exists for early-DNR utilization making it difficult to fully interpret the quantitative findings of this study. However, the purpose of the study was to explore potential differences in early-DNR utilization as a potential proxy for how health care systems respond to high strain situations in which both patients and physicians experience significant fear.
We speculate that several factors may have contributed to the slightly higher rates of early-DNR utilization during COVID-19 surge months compared with nonsurge months. The first factor may be that higher early-DNR utilization may be an honest representation of patient preferences. During COVID-19 surges, media channels were awash with reports of hospitals overflowing and incredibly high morbidity and mortality for patients, particularly older adults, who received MV.1–5 These reports, some of which presented misleading information, may have influenced patient opinions around resuscitative efforts and intubation as DNR orders often implies DNI. Multiple studies have demonstrated a growing need and role for palliative care practices in Emergency Departments (EDs) and some have suggested that there was increased utilization of palliative care practices in EDs during the COVID-19 pandemic.39–42 However, it is unclear why the fear of poor outcomes would fluctuate throughout the year from surge periods to nonsurge periods as sensational media reports continued for all of 2020.
It is possible that during surge periods medical teams were more likely to ask older adults about their care preferences. Some hospitals deployed palliative care teams to EDs and to ICUs during surges. These efforts may have led to more accurate assessment of goals of care but raises some questions about why differences only existed for older adults (i.e., did these teams only approach older adults and not younger adults) or whether these efforts highlight underutilization of DNR orders during nonsurge periods. Moreover, it also highlights the critical roles that palliative care teams can play during surge situations.
An alternative explanation for slightly higher early-DNR utilization only among older adults in 2020 is increased pressure on older adults as opposed to younger populations during surge periods. Fears on rationing of ventilators were prevalent throughout ICUs in 2020 with many hospitals and states updating their CSC rationing plans. Based on concerns that older adults might be pressured into accepting DNR orders, OCR issued specific direction that older patients could not be pressured or coerced into DNR orders regardless of activation of CSC plans.15,16 The National Academy of Medicine has also held that factors such as age alone should not be used to make triage decisions during times of crisis.7,8,43,44 Despite these guidelines, there were significant concerns that blanket DNR orders would be imposed on older adults during COVID-19 surges to preserve critical care resources.11,13 This study could be interpreted to suggest that older adults in California may have been subject to increased pressure to accept DNR orders at the time of admission as the increased DNR utilization was only seen during COVID-19 surge months when fears about rationing of care were at their highest.
Further evidence to support the possibility of increased pressure on older adults is the finding that no difference in early-DNR utilization was observed for younger adults and no seasonal variation was seen among older adults in 2019.
Increased DNR utilization had significant clinical implications. Similar to previous studies, we observed a nearly four-fold increase in mortality for patients with an early-DNR order.45–47 While a DNR order indicates a refusal of resuscitate efforts only (and in some cases intubation), the interpretation of DNR orders varies widely. Patients with DNR orders are less likely to receive multiple other forms of treatment ranging from central line placement to simple things like blood cultures.48–50 Therefore, even the small increase in early-DNR utilization in this study has clinical impact on patients that may or may not represent their true wishes.
There are several important limitations to this study. First, observed differences in early-DNR utilization were small in absolute terms and unmeasured confounding could affect the conclusions. However, the difference persisted across multiple sensitivity analyses, difference-in-difference analyses, and alternative cohort definitions suggesting that the differences was less likely to be affected by unmeasured confounding. The key observation that no difference in DNR utilization existed between younger adults in 2020 and older adults in 2019 highlights that the small difference among older adults in 2020 is a true observation. While unmeasured confounders may have impacted the regression analyses, the IPTW approach created a quasi-randomized situation further supporting the findings. Early-DNR status is more indicative of comorbid status and comorbidities can be identified in discharge datasets with high accuracy.30–32
Additionally, this study only included data from California, which may limit generalizability. We speculate that larger differences may be observed in regions with greater baseline variability in early-DNR utilization such as the South.35,51 This study also focused on DNR orders placed within the 24 hours of admission. It is possible that during surges, order entry may have been delayed due to clinicians being so much busier. In such a setting, an evaluation of DNR orders placed in the first 48 or 72 hours may have yielded greater variation in DNR utilization. The primary exposure was surge-month and different studies have defined COVID-19 surges differently. We used a combination of state-wide COVID-19 cases and hospitalizations to best estimate the impact that surges had on the psyche of both patients and providers. Alternative definitions of surges based on week rather than month or that use different definition for surges may arrive as slightly different conclusions. Rates of early-DNR orders were also higher in this study than previously reported. This may reflect the dual method of identification of early-DNR orders (unique identifier in the PDD and billing codes).
Importantly, our findings persisted when we restricted our analysis to only the unique early-DNR identifier in the PDD. Finally, we speculate about potential drivers of differences in DNR utilization among older adults but it is not possible to identify the main driver of our findings.
In this study, utilization of early-DNR orders among older adults during COVID-19 surges in 2020 was investigated. Increased early-DNR utilization among older adults during COVID-19 surges compared with nonsurge periods. The increase in early-DNR utilization was restricted to older adults in 2020 and not indicative of differences among younger patients or annual seasonal variation. The increase in early-DNR utilization had significant clinical implications with greater mortality for patients with early-DNR orders. It is unclear what contributed to these differences but possibilities include increased patient fear of critical care interventions based on widespread media reports as well as the potential for increased pressure on older patients to accept early-DNR orders during surge periods based on fears of exhausting resources. This study raises the potential impact of overstated clinical and media reports as well as the possibility of age-related bias during times of crisis. Future studies are needed to explore drivers of the differences observed in this study.
Authors' Contributions
A.B.M. and M.K.W. conceived the study. A.B.M. was responsible for data acquisition and analysis. A.B.M., I.S.D., C.B., and M.K.W. were responsible for data interpretation. A.B.M. drafted the article. A.B.M., I.S.D., C.B., and M.K.W. were responsible for critical revisions of the article. A.B.M. had full access to all the data in the study and takes responsibility for the integrity of the data and accuracy of the data analysis.
Ethics Approval
The work was approved by both the Colorado Multiple Institutional Review Board and the California Committee for the Protection of Human Subjects. Informed consent was waived.
Supplementary Material
Disclaimer
Contents are the authors' sole responsibility and do not necessarily represent official NIH views. The funders did not have any role in design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the article; or the decision to submit the article for publication. A.B.M. had full access to the data and takes full responsibility for the contents of this article.
Funding Information
A.B.M. is supported by NIH K23HL141704 and I.S.D. is supported by NIH R01NR016459. Additionally, the work was supported by NIH/NCATS Colorado CTSI grant no. UL1 TR002535.
Author Disclosure Statement
No competing financial interests exist.
Supplementary Material
References
- 1. Richardson S, Hirsch JS, Narasimhan M, et al. Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area. JAMA 2020;323(20):2052–2059; doi: 10.1001/jama.2020.6775 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Clarification of mortality rate and data in abstract, results, and table 2. JAMA 2020;323(20):2098; doi: 10.1001/jama.2020.7681 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bostock B. Video shows Italian army trucks transporting coffins from Italy's worst-hit city to remote cremation sites because morgues can't cope with more coronavirus deaths. New York, NY; 2020. Available from: https://www.businessinsider.com/coronavirus-italy-army-transport-coffins-bergamo-morgue-crisis-video-2020-3 [Last accessed: November 15].
- 4. Taylor A. An Unimaginable Toll. Washington D.C.; 2020. Available from: https://www.theatlantic.com/photo/2020/04/coronavirus-unimaginable-toll-photos/609652/ [Last accessed: November 15].
- 5. Voytko L. NYC Could Start ‘Temporary Interment’ For The Dead In Public Park, City Council Member Says; Mayor Denies. New York, NY; 2020. Available from: https://www.forbes.com/sites/lisettevoytko/2020/04/06/nyc-likely-to-start-temporary-interment-for-the-dead-in-public-park-city-councillor-says/?sh=5e8117db1799 [Last accessed: November 15].
- 6. Cleveland Manchanda EC, Sanky C, Appel JM. Crisis standards of care in the USA: A systematic review and implications for equity amidst COVID-19. J Racial Ethn Health Disparities 2021;8(4):824–836; doi: 10.1007/s40615-020-00840-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Institute of Medicine. Crisis Standards of Care: A Systems Framework for Catastrophic Disaster Response. The National Academies Press: Washington, DC, USA; 2012. [PubMed] [Google Scholar]
- 8. Institute of Medicine. Guidance for Establishing Crisis Standards of Care for Use in Disaster Situations: A Letter Report. The National Academies Press: Washington, DC, USA; 2009. [PubMed] [Google Scholar]
- 9. OCR Reaches Early Case Resolution With Alabama After It Removes Discriminatory Ventilator Triaging Guidelines. Washington DC; 2020. Available from: https://www.hhs.gov/about/news/2020/04/08/ocr-reaches-early-case-resolution-alabama-after-it-removes-discriminatory-ventilator-triaging.html [Last accessed: July 1].
- 10. Alexiou G. Doctors Issuing Unlawful ‘Do Not Resuscitate’ Orders For Disabled Covid Patients ‘Outrageous’. Jersey City, NJ; 2020. Available from: https://www.forbes.com/sites/gusalexiou/2020/06/23/unlawful-do-not-resuscitate-orders-for-disabled-covid-patients-outrageous [Last accessed: January 15].
- 11. Cha AE. Hospitals Consider Universal Do-Not-Resuscitate Orders for Coronavirus Patients. The Washington Post: Washington, DC, USA; 2020. [Google Scholar]
- 12. Howard J, Bruer W, Christensen J. Hospitals consider changes to do-not-resuscitate situations amid coronavirus pandemic. Atlanta, GA; 2020. Available from: https://www.cnn.com/2020/03/26/health/coronavirus-hospitals-do-not-resuscitate-bn/index.html [Last accessed: August 12].
- 13. Bledsoe TA, Jokela JA, Deep NN, et al. Universal do-not-resuscitate orders, social worth, and life-years: Opposing discriminatory approaches to the allocation of resources during the COVID-19 pandemic and other health system catastrophes. Ann Intern Med 2020;173(3):230–232; doi: 10.7326/m20-1862 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Civil Rights Protections Prohibiting Race, Color and National Origin Discrimination During COVID-19. Washington DC; 2020. Available from: https://www.hhs.gov/sites/default/files/title-vi-bulletin.pdf [Last accessed: November 21].
- 15. Civil Rights, HIPAA, and the Coronavirus Disease 2019 (COVID-19). Washington DC; 2020. Available from: https://www.hhs.gov/sites/default/files/ocr-bulletin-3-28-20.pdf [Last accessed: November 15].
- 16. Civil Rights and COVID-19. Washington DC; 2022. Available from: https://www.hhs.gov/civil-rights/for-providers/civil-rights-covid19/index.html [Last accessed: March 15].
- 17. Yang W, Kandula S, Huynh M, et al. Estimating the infection-fatality risk of SARS-CoV-2 in New York City during the spring 2020 pandemic wave: A model-based analysis. Lancet Infect Dis 2021;21(2):203–212; doi: 10.1016/s1473-3099(20)30769-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Horwitz LI, Jones SA, Cerfolio RJ, et al. Trends in COVID-19 risk-adjusted mortality rates. J Hosp Med 2021;16(2):90–92; doi: 10.12788/jhm.3552 [DOI] [PubMed] [Google Scholar]
- 19. Roth GA, Emmons-Bell S, Alger HM, et al. Trends in patient characteristics and COVID-19 in-hospital mortality in the United States during the COVID-19 pandemic. JAMA Netw Open 2021;4(5):e218828; doi: 10.1001/jamanetworkopen.2021.8828 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Mehta AB, Wynia MK. Crisis standards of care-more than just a thought experiment? Hastings Cent Rep 2021;51(5):53–55; doi: 10.1002/hast.1288 [DOI] [PubMed] [Google Scholar]
- 21. Shapiro J. Oregon Hospitals Didn't Have Shortages. So Why Were Disabled People Denied Care? Washington DC; 2020. Available from: https://www.npr.org/2020/12/21/946292119/oregon-hospitals-didnt-have-shortages-so-why-were-disabled-people-denied-care [Last accessed: July 20].
- 22. Shapiro J. One Man's COVID-19 Death Raises The Worst Fears Of Many People With Disabilities. Washington DC; 2020. Available from: https://www.npr.org/2020/07/31/896882268/one-mans-covid-19-death-raises-the-worst-fears-of-many-people-with-disabilities [Last accessed: January 3].
- 23. Research Data Request Information. Sacramento, CA; 2022. Available from: https://hcai.ca.gov/data-and-reports/research-data-request-information/ [Last accessed: April 2].
- 24. Data and Reports. Sacramento, CA; 2022. Available from: https://hcai.ca.gov/data-and-reports/ [Last accessed: April 1].
- 25. Goldman LE, Chu PW, Osmond D, et al. Accuracy of do not resuscitate (DNR) in administrative data. Medical care research and review: MCRR 2013;70(1):98–112; doi: 10.1177/1077558712458455 [DOI] [PubMed] [Google Scholar]
- 26. Goldman LE, Chu PW, Prothro C, et al. Accuracy of condition present on admission, do not resuscitate, and e-codes in California Patient Discharge Data: Prepared for the Office of Statewide Health Planning and Development, Healthcare Outcomes Center. 2011. Available from: https://www.oshpd.ca.gov/documents/HID/PDDValidation/PDD_Validation_Study.pdf [Last accessed: March 3].
- 27. Procedure Classes Refined for ICD-10-PCS. Rockville, MD; 2022. Available from: https://www.hcup-us.ahrq.gov/toolssoftware/procedureicd10/procedure_icd10.jsp [Last accessed: October 15].
- 28. California Open Data Portal: COVID-19 Hospital Data. Sacramento, CA; 2022. Available from: https://data.ca.gov/dataset/covid-19-hospital-data1 [Last accessed: October 1].
- 29. Pollock BD, Herrin J, Neville MR, et al. Association of do-not-resuscitate patient case mix with publicly reported risk-standardized hospital mortality and readmission rates. JAMA Netw Open 2020;3(7):e2010383; doi: 10.1001/jamanetworkopen.2020.10383 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Marrie TJ, Fine MJ, Kapoor WN, et al. Community-acquired pneumonia and do not resuscitate orders. J Am Geriatr Soc 2002;50(2):290–299; doi: jgs50061 [pii] [DOI] [PubMed] [Google Scholar]
- 31. Elixhauser Comorbidity Software Refined for ICD-10-CM. Rockville, MD; 2021. Available from: https://www.hcup-us.ahrq.gov/toolssoftware/comorbidityicd10/comorbidity_icd10.jsp [Last accessed: February 15].
- 32. Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care 1998;36(1):8–27. [DOI] [PubMed] [Google Scholar]
- 33. Bosch NA, Law AC, Rucci JM, et al. Predictive validity of the sequential organ failure assessment score versus claims-based scores among critically ill patients. Ann Am Thorac Soc 2022; doi: 10.1513/AnnalsATS.202111-1251RL [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Hakim RB, Teno JM, Harrell FEJr., et al. Factors associated with do-not-resuscitate orders: Patients' preferences, prognoses, and physicians' judgments. SUPPORT Investigators. Study to understand prognoses and preferences for outcomes and risks of treatment. Ann Intern Med 1996;125(4):284–293. [DOI] [PubMed] [Google Scholar]
- 35. Zingmond DS, Wenger NS. Regional and institutional variation in the initiation of early do-not-resuscitate orders. Arch Intern Med 2005;165(15):1705–1712; doi: 165/15/1705 [pii] [DOI] [PubMed] [Google Scholar]
- 36. Shepardson LB, Gordon HS, Ibrahim SA, et al. Racial variation in the use of do-not-resuscitate orders. J Gen Intern Med 1999;14(1):15–20; doi: 10.1046/j.1525-1497.1999.00275.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Zheng NT, Mukamel DB, Caprio T, et al. Racial disparities in in-hospital death and hospice use among nursing home residents at the end of life. Med Care 2011;49(11):992–998; doi: 10.1097/MLR.0b013e318236384e [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Steyerberg EW. Clinical Prediction Models. Springer: Cham, Switzerland; 2009. [Google Scholar]
- 39. Bayuo J, Agbeko AE, Acheampong EK, et al. Palliative care interventions for adults in the emergency department: A review of components, delivery models, and outcomes. Acad Emerg Med 2022;29(11):1357–1378; doi: 10.1111/acem.14508 [DOI] [PubMed] [Google Scholar]
- 40. Eygnor JK, Rosenau AM, Burmeister DB, et al. Palliative care in the emergency department during a COVID-19 pandemic. Am J Emerg Med 2021;45:516–518; doi: 10.1016/j.ajem.2020.07.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Long DA, Koyfman A, Long B. Oncologic emergencies: palliative care in the emergency department setting. J Emerg Med 2021;60(2):175–191; doi: 10.1016/j.jemermed.2020.09.027 [DOI] [PubMed] [Google Scholar]
- 42. Tan A, Durbin M, Chung FR, et al. Design and implementation of a clinical decision support tool for primary palliative Care for Emergency Medicine (PRIM-ER). BMC Med Inform Decis Mak 2020;20(1):13; doi: 10.1186/s12911-020-1021-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Hick JL, Hanfling D, Wynia MK, et al. Crisis Standards of Care and COVID-19: What Did We Learn? How Do We Ensure Equity? What Should We Do? National Academy of Medicine: Washington, DC; 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. National Academies of Sciences Engineering and Medicine. In: Rapid Expert Consultations on the COVID-19 Pandemic: March 14, 2020–April 8, 2020. National Academies Press (US) Copyright 2020 by the National Academy of Sciences. All rights reserved: Washington, DC, USa; 2020. [PubMed] [Google Scholar]
- 45. Shepardson LB, Youngner SJ, Speroff T, et al. Increased risk of death in patients with do-not-resuscitate orders. Med Care 1999;37(8):727–737. [DOI] [PubMed] [Google Scholar]
- 46. Tabak YP, Johannes RS, Silber JH, et al. Should Do-Not-Resuscitate status be included as a mortality risk adjustor? The impact of DNR variations on performance reporting. Med Care 2005;43(7):658–666; doi: 00005650-200507000-00004 [pii] [DOI] [PubMed] [Google Scholar]
- 47. Walkey AJ, Weinberg J, Wiener RS, et al. Association of do-not-resuscitate orders and hospital mortality rate among patients with pneumonia. JAMA Intern Med 2016;176(1):97–104; doi: 10.1001/jamainternmed.2015.6324 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. La Puma J, Silverstein MD, Stocking CB, et al. Life-sustaining treatment. A prospective study of patients with DNR orders in a teaching hospital. Arch Intern Med 1988;148(10):2193–2198. [DOI] [PubMed] [Google Scholar]
- 49. Stevenson EK, Mehter HM, Walkey AJ, et al. Association between do not resuscitate/do not intubate status and resident physician decision-making. A national survey. Ann Am Thorac Soc 2017;14(4):536–542; doi: 10.1513/AnnalsATS.201610-798OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Walkey AJ, Weinberg J, Wiener RS, et al. Hospital variation in utilization of life-sustaining treatments among patients with do not resuscitate orders. Health Serv Res 2017; doi: 10.1111/1475-6773.12651 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Nicholas LH, Langa KM, Iwashyna TJ, et al. Regional variation in the association between advance directives and end-of-life Medicare expenditures. JAMA 2011;306(13):1447–1453; doi: 10.1001/jama.2011.1410 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.