Using pooled electronic health records data to conduct pharmacoepidemiology safety studies: Challenges and lessons learned.

PURPOSE: We assessed the suitability of pooled electronic health record (EHR) data from clinical research networks (CRNs) of the patient-centered outcomes research network to conduct studies of the association between tumor necrosis factor inhibitors (TNFi) and infections. METHODS: EHR data from patients with one of seven autoimmune diseases were obtained from three CRNs and pooled. Person-level linkage of CRN data and Centers for Medicare and Medicaid Services (CMS) fee-for-service claims data was performed where possible. Using filled prescriptions from CMS claims data as the gold standard, we assessed the misclassification of EHR-based new (incident) user definitions. Among new users of TNFi, we assessed subsequent rates of hospitalized infection in EHR and CMS data. RESULTS: The study included 45 483 new users of TNFi, of whom 1416 were successfully linked to their CMS claims. Overall, 44% of new EHR TNFi prescriptions were not associated with medication claims. Our most specific new user definition had a misclassification rate of 3.5%-16.4% for prevalent use, depending on the medication. Greater than 80% of CRN prescriptions had either zero refills or missing refill data. Compared to using EHR data alone, there was a 2- to 8-fold increase in hospitalized infection rates when CMS claims data were added to the analysis. CONCLUSIONS: EHR data substantially misclassified TNFi exposure and underestimated the incidence of hospitalized infections compared to claims data. EHR-based new user definitions were reasonably accurate. Overall, using CRN data for pharmacoepidemiology studies is challenging, especially for biologics, and would benefit from supplementation by other sources.

Development and Validation of an Emergency Department Electronic Medical Record Gout Flare Alert.

OBJECTIVE: Patients with acute gout are frequently treated in the emergency department (ED) and represent a typically underresourced and understudied population. A key limitation for gout research in the ED is the timely ability to identify acute gout patients. Our goal was to refine a multicriteria, electronic medical record alert for gout flares and to determine its diagnostic characteristics in the ED. METHODS: The gout flare alert used electronic medical record data from ED nursing notes and was triggered by the term 'gout' preceding past medical history in the chief complaint, the term 'gout' and a musculoskeletal problem in the chief complaint, or the term 'gout' in the problem list and a musculoskeletal chief complaint. We validated its diagnostic properties to assess presence/absence of gout through manual medical record review using adjudicated expert consensus as the gold standard. RESULTS: In January 2020, we analyzed 202 patient records from 2 university-based EDs; from these records, 57 patients were identified by our gout flare alert, and 145 were identified by other means as potentially having an acute gout flare. The gout flare alert's positive predictive value was 47% (95% confidence interval [95% CI] 34-60%), negative predictive value was 94% (95% CI 90-98%), sensitivity was 75% (95% CI 61-89%), and specificity was 82% (95% CI 76-88%). The diagnostic properties were similar at both institutions. CONCLUSION: Our multicomponent gout flare alert had reasonable sensitivity and specificity, albeit a modest positive predictive value. An electronic gout flare alert may help enable the conduct of gout research in the ED setting.

Anti-tRNA synthetase syndrome interstitial lung disease: A single center experience.

BACKGROUND: Recognition of Anti-tRNA synthetase (ARS) related interstitial lung disease (ILD) is key to ensuring patients have prompt access to immunosuppressive therapies. The purpose of this retrospective cohort study was to identify factors that may delay recognition of ARS-ILD. METHODS: Patients seen at Vanderbilt University Medical Center between 9/17/2017-10/31/2018 were included in this observational cohort. Clinical and laboratory features were obtained via chart abstraction. Kruskal-Wallis ANOVA, Mann-Whitney U, and Fisher's exact t tests were utilized to determine statistical significance. RESULTS: Patients with ARS were found to have ILD in 51.9% of cases, which was comparable to the frequency of ILD in systemic sclerosis (59.5%). The severity of FVC reduction in ARS (53.2%) was comparable to diffuse cutaneous systemic sclerosis (56.8%, p = 0.48) and greater than dermatomyositis (66.9%, p = 0.005) or limited cutaneous systemic sclerosis (71.8%, p = 0.005). Frank honeycombing was seen with ARS antibodies but not other myositis autoantibodies. ARS patients were more likely to first present to a pulmonary provider in a tertiary care setting (53.6%), likely due to fewer extrapulmonary manifestations. Only 33% of ARS-ILD were anti-nuclear antibody, rheumatoid factor, or anti-cyclic citrullinated peptide positive. Patients with ARS-ILD had a two-fold longer median time to diagnosis compared to other myositis-ILD patients (11.0 months, IQR 8.5-43 months vs. 5.0 months, IQR 3.0-9.0 months, p = 0.003). CONCLUSIONS: ARS patients without prominent extra-pulmonary manifestations are at high risk for not being recognized as having a connective tissue disease related ILD and miscategorized as usual interstitial pneumonia/idiopathic pulmonary fibrosis without comprehensive serologies.

CD19 + CD21lo/neg cells are increased in systemic sclerosis-associated interstitial lung disease.

Interstitial lung disease (ILD) represents a significant cause of morbidity and mortality in systemic sclerosis (SSc). The purpose of this study was to examine recirculating lymphocytes from SSc patients for potential biomarkers of interstitial lung disease (ILD). Peripheral blood mononuclear cells (PBMCs) were isolated from patients with SSc and healthy controls enrolled in the Vanderbilt University Myositis and Scleroderma Treatment Initiative Center cohort between 9/2017-6/2019. Clinical phenotyping was performed by chart abstraction. Immunophenotyping was performed using both mass cytometry and fluorescence cytometry combined with t-distributed stochastic neighbor embedding analysis and traditional biaxial gating. This study included 34 patients with SSc-ILD, 14 patients without SSc-ILD, and 25 healthy controls. CD21lo/neg cells are significantly increased in SSc-ILD but not in SSc without ILD (15.4 ± 13.3% vs. 5.8 ± 0.9%, p = 0.002) or healthy controls (5.0 ± 0.5%, p < 0.0001). While CD21lo/neg B cells can be identified from a single biaxial gate, tSNE analysis reveals that the biaxial gate is comprised of multiple distinct subsets, all of which are increased in SSc-ILD. CD21lo/neg cells in both healthy controls and SSc-ILD are predominantly tBET positive and do not have intracellular CD21. Immunohistochemistry staining demonstrated that CD21lo/neg B cells diffusely infiltrate the lung parenchyma of an SSc-ILD patient. Additional work is needed to validate this biomarker in larger cohorts and longitudinal studies and to understand the role of these cells in SSc-ILD.

Thirty-Day Hospital Readmissions for Acute Gout in the United States: Analysis of the Nationwide Readmission Database.

BACKGROUND/OBJECTIVE: The aims of this study were to describe the rates and characteristics of nonelective 30-day readmission among adult patients hospitalized for acute gout and to assess predictors of readmission. METHODS: We analyzed the 2017 Nationwide Readmission Database. Gout hospitalizations were identified using the International Classification of Diseases, Tenth Revision, Clinical Modification code. Hospitalizations for adult patients were included. We excluded planned or elective readmissions. We utilized χ2 tests to compare baseline characteristics between readmissions and index hospitalizations. We used multivariate Cox regression to identify independent predictors of readmissions. RESULTS: A total of 11,727 index adult hospitalizations with acute gout listed as the principal diagnosis were discharged alive and included. One thousand five hundred ninety-four (13.6%) readmissions occurred within 30 days. Acute gout was the most common reason for readmission. Readmissions had higher inpatient mortality (2.4% vs 0.1%, p < 0.0001), greater mean age (68.1 vs 67.0 years, p = 0.021), and longer hospital length of stay (5.9 vs 3.8 days, p < 0.0001) compared with index hospitalizations. Charlson Comorbidity Index scores of ≥2 (score 2: adjusted hazards ratio [AHR], 1.67; p = 0.001; score ≥3: AHR, 2.08; p < 0.0001), APR-DRG (All Patients Refined Diagnosis Related Groups) severity levels ≥2 (level 2: AHR, 1.43; p = 0.044; level 3: AHR, 1.83; p = 0.002; level 4: AHR, 2.38; p = 0.002), admission to metropolitan hospital (AHR, 1.83; p = 0.012), atrial fibrillation (AHR, 1.31; p = 0.004), and anemia (AHR, 1.30; p = 0.001) were significantly associated with 30-day readmissions. CONCLUSIONS: Acute gout readmissions were associated with worse outcomes compared with index hospitalizations. Charlson Comorbidity Index scores ≥2, APR-DRG severity levels ≥2, admission to metropolitan hospital, atrial fibrillation, and anemia were significant predictors of readmission.

Patient-Reported Outcomes in Lupus.

Patient-reported outcome (PRO) was identified as a core systemic lupus erythematosus (SLE) outcome in 1999. More than 20 years later, however, generic PRO measures evaluating impact in SLE are used mainly for research. Generic and disease-targeted PRO tools have unique advantages. Significant progress in identification of patient disease-relevant PRO concepts and development of new PRO tools for SLE has occurred over the past 20 years. Further research needs to focus on responsiveness and minimally important differences of existing, promising PRO tools to facilitate their use in SLE patient care and research.

A Case of Acute Neck Pain: The Crowned Dens Syndrome.

Crowned dens syndrome (CDS) is a relatively uncommon presentation of calcium pyrophosphate dihydrate (CPPD) deposition disease that manifests as acute attacks of neck pain with fever, neck rigidity and elevated inflammatory markers related to radiodense deposits of CPPD in ligaments around the odontoid process. We present a case of CDS.

National Landscape of Unplanned 30-Day Readmissions in Patients With Left Ventricular Assist Device Implantation.

The number of patients with advanced heart failure receiving left ventricular assist device (LVAD) implantation has increased dramatically over the last decade. There are limited data available about the nationwide trends of complications leading to readmissions after implantation of contemporary devices. Patients who underwent LVAD implantation from January 2013 to December 2013 were identified using International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) code 37.66 from the Healthcare Cost and Utilization Project's National Readmission Database. The top causes of unplanned 30-day readmission after LVAD implantation were determined. Survey logistic regression was used to analyze the significant predictors of readmission. In 2013, there were 2,235 patients with an LVAD implantation. Of them, 665 (29.7%) had at least 1 unplanned readmission within 30 days, out of which 289 (43.4%) occurred within 10 days after discharge. Implant complications (14.9%), congestive heart failure (11.7%), and gastrointestinal bleeding (8.4%) were the top 3 diagnoses for the first readmission and accounted for more than a third of all readmissions. Significant predictors of readmissions included a prolonged length of stay during the index admission, Medicare insurance, and discharge to short-term facility. In conclusion, despite increased experience with LVADs, unplanned readmissions within 30 days of implantation remain significantly high.

National Trends and Outcomes in Dialysis-Requiring Acute Kidney Injury in Heart Failure: 2002-2013.

BACKGROUND: Dialysis-requiring acute kidney injury (D-AKI) is a serious complication in hospitalized heart failure (HF) patients. However, data on national trends are lacking after 2002. METHODS: We used the Nationwide Inpatient Sample (2002-2013) to identify HF hospitalizations with and without D-AKI. We analyzed trends in incidence, in-hospital mortality, length of stay (LoS), and cost. We calculated adjusted odds ratios (aORs) for predictors of D-AKI and for outcomes including in-hospital mortality and adverse discharge (discharge to skilled nursing facilities, nursing homes, etc). RESULTS: We identified 11,205,743 HF hospitalizations. Across 2002-2013, the incidence of D-AKI doubled from 0.51% to 1.09%. We found male sex, younger age, African-American and Hispanic race, and various comorbidities and procedures, such as sepsis and mechanical ventilation, to be independent predictors of D-AKI in HF hospitalizations. D-AKI was associated with higher odds of in-hospital mortality (aOR 2.49, 95% confidence interval [CI] 2.36-2.63; P < .01) and adverse discharge (aOR 2.04, 95% CI 1.95-2.13; P < .01). In-hospital mortality and attributable risk of mortality due to D-AKI decreased across 2002-2013. LoS and cost also decreased across this period. CONCLUSIONS: The incidence of D-AKI in HF hospitalizations doubled across 2002-2013. Despite declining in-hospital mortality, LoS, and cost, D-AKI was associated with worse outcomes.

Nationwide Trends in Hospital Outcomes and Utilization After Lower Limb Revascularization in Patients on Hemodialysis.

OBJECTIVES: This study aimed to describe the temporal trends and outcomes of endovascular and surgical revascularization in a large, nationally representative sample of patients with end-stage renal disease on hemodialysis hospitalized for peripheral artery disease (PAD). BACKGROUND: PAD is prevalent among patients with end-stage renal disease on hemodialysis and is associated with significant morbidity and mortality. There is a paucity of information on trends in endovascular and surgical revascularization and post-procedure outcomes in this population. METHODS: We used the Nationwide Inpatient Sample (2002 to 2012) to identify hemodialysis patients undergoing endovascular or surgical procedures for PAD using diagnostic and procedural codes. We compared trends in amputation, post-procedure complications, mortality, length of stay, and costs between the 2 groups using trend tests and logistic regression. RESULTS: There were 77,049 endovascular and 29,556 surgical procedures for PAD in hemodialysis patients. Trend analysis showed that endovascular procedures increased by nearly 3-fold, whereas there was a reciprocal decrease in surgical revascularization. Post-procedure complication rates were relatively stable in persons undergoing endovascular procedures but nearly doubled in those undergoing surgery. Surgery was associated with 1.8 times adjusted odds (95% confidence interval: 1.60 to 2.02) for complications and 1.6 times the adjusted odds for amputations (95% confidence interval: 1.40 to 1.75) but had similar mortality (adjusted odds ratio: 1.05; 95% confidence interval: 0.85 to 1.29) compared with endovascular procedures. Length of stay for endovascular procedures remained stable, whereas a decrease was seen for surgical procedures. Overall costs increased marginally for both procedures. CONCLUSIONS: Rates of endovascular procedures have increased, whereas those of surgeries have decreased. Surgical revascularization is associated with higher odds of overall complications. Further prospective studies and clinical trials are required to analyze the relationship between the severity of PAD and the revascularization strategy chosen.

Impact of dialysis requirement on outcomes in tumor lysis syndrome.

Tumor lysis syndrome (TLS) is a life threatening emergency due to destruction and massive release of intracellular metabolites from cancer cells often resulting in acute kidney injury (AKI), sometimes severe enough to require dialysis (AKI-D). The impact of dialysis requirement in AKI has not been explored. We utilized data from the Nationwide Inpatient Sample and using International Classification of Diseases, 9th Revision, diagnoses codes for TLS, AKI and dialysis, evaluated the incidence, risk factors and impact of AKI-D on mortality, adverse discharge and length of stay (LOS). Survey multivariable logistic regression was used to compute adjusted Odds Ratios (aOR and 95% confidence intervals (CI). An estimated 12% (2,919) of all TLS hospitalizations (n = 22 875) develop AK-D. After adjustment for confounders, AKI-D was associated with greater odds of mortality (aOR 1.98; (95% CI 1.60-2.45)), adverse discharge (aOR 1.63 (95% CI 1.19-2.24)) and longer LOS (19 vs 14.6 days; P < 0.01) compared with those without AKI-D. Further studies to evaluate the association of AKI-D on long-term outcomes in patients with TLS are needed.

National Trends and Impact of Acute Kidney Injury Requiring Hemodialysis in Hospitalizations With Atrial Fibrillation.

BACKGROUND: Atrial fibrillation (AF) is a common cause for hospitalization, but there are limited data regarding acute kidney injury requiring dialysis (AKI-D) in AF hospitalizations. We aimed to assess temporal trends and outcomes in AF hospitalizations complicated by AKI-D utilizing a nationally representative database. METHODS AND RESULTS: Utilizing the Nationwide Inpatient Sample, AF hospitalizations and AKI-D were identified using diagnostic and procedure codes. Trends were analyzed overall and within subgroups and utilized multivariable logistic regression to generate adjusted odds ratios (aOR) for predictors and outcomes including mortality and adverse discharge. Between 2003 and 2012, 3751 (0.11%) of 3 497 677 AF hospitalizations were complicated by AKI-D. The trend increased from 0.3/1000 hospitalizations in 2003 to 1.5/1000 hospitalizations in 2012, with higher increases in males and black patients. Temporal changes in demographics and comorbidities explained a substantial proportion but not the entire trend. Significant comorbidities associated with AKI-D included mechanical ventilation (aOR 13.12; 95% CI 9.88-17.43); sepsis (aOR 8.20; 95% CI 6.00-11.20); and liver failure (aOR 3.72; 95% CI 2.92-4.75). AKI-D was associated with higher risk of in-hospital mortality (aOR 3.54; 95% CI 2.81-4.47) and adverse discharge (aOR 4.01; 95% CI 3.12-5.17). Although percentage mortality within AKI-D decreased over the decade, attributable risk percentage mortality remained stable. CONCLUSIONS: AF hospitalizations complicated by AKI-D have quintupled over the last decade with differential increase by demographic groups. AKI-D is associated with significant morbidity and mortality. Without effective AKI-D therapies, focus should be on early risk stratification and prevention to avoid this devastating complication.

Development and validation of case-finding algorithms for the identification of patients with anti-neutrophil cytoplasmic antibody-associated vasculitis in large healthcare administrative databases.

PURPOSE: The aim of this study was to develop and validate case-finding algorithms for granulomatosis with polyangiitis (Wegener's, GPA), microscopic polyangiitis (MPA), and eosinophilic GPA (Churg-Strauss, EGPA). METHODS: Two hundred fifty patients per disease were randomly selected from two large healthcare systems using the International Classification of Diseases version 9 (ICD9) codes for GPA/EGPA (446.4) and MPA (446.0). Sixteen case-finding algorithms were constructed using a combination of ICD9 code, encounter type (inpatient or outpatient), physician specialty, use of immunosuppressive medications, and the anti-neutrophil cytoplasmic antibody type. Algorithms with the highest average positive predictive value (PPV) were validated in a third healthcare system. RESULTS: An algorithm excluding patients with eosinophilia or asthma and including the encounter type and physician specialty had the highest PPV for GPA (92.4%). An algorithm including patients with eosinophilia and asthma and the physician specialty had the highest PPV for EGPA (100%). An algorithm including patients with one of the diagnoses (alveolar hemorrhage, interstitial lung disease, glomerulonephritis, and acute or chronic kidney disease), encounter type, physician specialty, and immunosuppressive medications had the highest PPV for MPA (76.2%). When validated in a third healthcare system, these algorithms had high PPV (85.9% for GPA, 85.7% for EGPA, and 61.5% for MPA). Adding the anti-neutrophil cytoplasmic antibody type increased the PPV to 94.4%, 100%, and 81.2% for GPA, EGPA, and MPA, respectively. CONCLUSION: Case-finding algorithms accurately identify patients with GPA, EGPA, and MPA in administrative databases. These algorithms can be used to assemble population-based cohorts and facilitate future research in epidemiology, drug safety, and comparative effectiveness. Copyright © 2016 John Wiley & Sons, Ltd.

Epidemiology of inpatient stay in Parkinson's disease in the United States: Insights from the Nationwide Inpatient Sample.

The total number of people living with Parkinson's disease (PD) worldwide is expected to double by 2030. The risk factors for emergency department visits in PD patients have been described before, however, there is limited data on inpatient hospitalizations of PD patients. We derived our study cohort from the Nationwide Inpatient Sample (NIS) database from 2002-2011. The NIS is a stratified 20% sample of discharges from all U.S. hospitals. We extracted causes of hospitalization using International Classification of Diseases, 9th revision, Clinical Modification (ICD-9-CM) codes and calculated inpatient mortality, length of stay and cost. Further, the significance of trends over 10 years was assessed. A total of 3,015,645 (weighted) admissions of PD patients were documented from 2002-2011. Pneumonia, urinary tract infection (UTI), septicemia and aspiration pneumonitis were the most common causes of admission, of which incidence of sepsis and UTI was trending up. Of all causes, 3.9% of the admissions resulted in inpatient mortality. Inpatient mortality for PD patients decreased from 4.9% in 2002 to 3.3% in 2011 (p<0.001). The median length of stay has also steadily declined from 3.6days in 2002 to 2.3days in 2011. However, the inflation-adjusted cost of care has been steadily rising, from $22,250 per hospitalization in 2002 to $37,942 in 2011. We conclude that the epidemiology of inpatient admissions in PD has changed significantly over the last decade. Our study underscores the need for future, in-depth prospective studies to explore this changing disease spectrum to design preventive measures and targeted interventions.

A Nationwide Analysis of Outcomes of Weekend Admissions for Intracerebral Hemorrhage Shows Disparities Based on Hospital Teaching Status.

BACKGROUND AND PURPOSE: With the "weekend effect" being well described, the Brain Attack Coalition released a set of "best practice" guidelines in 2005, with the goal to uniformly provide standard of care to patients with stroke. We attempted to define a "weekend effect" in outcomes among patients with intracranial hemorrhage (ICH) over the last decade, utilizing the Nationwide Inpatient Sample (NIS) data. We also attempted to analyze the trend of such an effect. MATERIALS AND METHODS: We determined the association of ICH weekend admissions with hospital outcomes including mortality, adverse discharge, length of stay, and cost compared to weekday admissions using multivariable logistic regression. We extracted our study cohort from the NIS, the largest all-payer data set in the United States. RESULTS: Of 485 329 ICH admissions from 2002 to 2011, 27.5% were weekend admissions. Overall, weekend admissions were associated with 11% higher odds of in-hospital mortality. When analyzed in 3-year groups, excess mortality of weekend admissions showed temporal decline. There was higher mortality with weekend admissions in nonteaching hospitals persisted (odds ratios 1.16, 1.13, and 1.09, respectively, for 3-year subgroups). Patients admitted during weekends were also 9% more likely to have an adverse discharge (odds ratio 1.09; 95% confidence interval: 1.07-1.11; P < .001) with no variation by hospital status. There was no effect of a weekend admission on either length of stay or cost of care. CONCLUSION: Nontraumatic ICH admissions on weekends have higher in-hospital mortality and adverse discharge. This demonstrates need for in-depth review for elucidating this discrepancy and stricter adherence to standard-of-care guidelines to ensure uniform care.

National trends of acute kidney injury requiring dialysis in decompensated cirrhosis hospitalizations in the United States.

BACKGROUND AND AIMS: Cirrhosis affects 5.5 million patients with estimated costs of US$4 billion. Previous studies about dialysis requiring acute kidney injury (AKI-D) in decompensated cirrhosis (DC) are from a single center/year. We aimed to describe national trends of incidence and impact of AKI-D in DC hospitalizations. METHODS: We extracted our cohort from the Nationwide Inpatient Sample (NIS) from 2006-2012. We identified hospitalizations with DC and AKI-D by validated ICD9 codes. We analyzed temporal changes in DC hospitalizations complicated by AKI-D and utilized multivariable logistic regression models to estimate AKI-D impact on hospital mortality. RESULTS: We identified a total of 3,655,700 adult DC hospitalizations from 2006 to 2012 of which 78,015 (2.1 %) had AKI-D. The proportion with AKI-D increased from 1.5 % in 2006 to 2.23 % in 2012; it was stable between 2009 and 2012 despite an increase in absolute numbers from 6773 to 13,930. The overall hospital mortality was significantly higher in hospitalizations with AKI-D versus those without (40.87 vs. 6.96 %; p < 0.001). In an adjusted multivariable analysis, adjusted odds ratio for mortality was 2.17 (95 % CI 2.06-2.28; p < 0.01) with AKI-D, which was stable from 2006 to 2012. Changes in demographics and increases in acute/chronic comorbidities and procedures explained temporal changes in AKI-D. CONCLUSIONS: Proportion of DC hospitalizations with AKI-D increased from 2006 to 2009, and although this was stable from 2009 to 2012, there was an increase in absolute cases. These results elucidate the burden of AKI-D on DC hospitalizations and excess associated mortality, as well as highlight the importance of prevention, early diagnosis and testing of novel interventions in this vulnerable population.

A National Perspective of Do-Not-Resuscitate Order Utilization Predictors in Intracerebral Hemorrhage.

Nontraumatic intracerebral hemorrhage (ICH) is associated with substantial morbidity and mortality. Do-not-resuscitate (DNR) orders are linked to poorer outcomes in patients with ICH, possibly due to less active management. Demographic, regional, and social factors, not related to ICH severity, have not been adequately looked at as significant predictors of DNR utilization. We reviewed the Healthcare Cost and Utilization Project's Nationwide Inpatient Sample (NIS) database in 2011 for adult ICH admissions and DNR status. We generated hierarchical 2-level multivariate regression models to estimate adjusted odds ratios. We analyzed 25 768 ICH hospitalizations, 18% of which (4620 hospitalizations) had DNR orders, corresponding to national estimates of 126 254 and 22 668, respectively. In multivariable regression, female gender, white or Hispanic/Latino ethnicity, no insurance coverage, and teaching hospitals were significantly associated with increased DNR utilization after adjusting for confounders. There was also significantly more interhospital variability in the lowest quartile of hospital volume. In conclusion, demographic factors and insurance status are significantly associated with increased DNR utilization, with more individual hospital variability in low-volume hospitals. The reasons for this are likely qualitative and linked to patient, provider, and hospital practices.