Should Do-Not-Resuscitate Status be Included as a Mortality Risk Adjustor?

Utilization of Hospital Do-Not-Resuscitate Orders in Older Adults During COVID-19 Surges in 2020

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.

Exploratory analysis of novel electronic health record variables for quantification of healthcare delivery strain, prediction of mortality, and prediction of imminent discharge

OBJECTIVE

To explore the relationship between novel, time-varying predictors for healthcare delivery strain (eg, counts of patient orders per hour) and imminent discharge and in-hospital mortality.

MATERIALS AND METHODS

We conducted a retrospective cohort study using data from adults hospitalized at 21 Kaiser Permanente Northern California hospitals between November 1, 2015 and October 31, 2020 and the nurses caring for them. Patient data extracted included demographics, diagnoses, severity measures, occupancy metrics, and process of care metrics (eg, counts of intravenous drip orders per hour). We linked these data to individual registered nurse records and created multiple dynamic, time-varying predictors (eg, mean acute severity of illness for all patients cared for by a nurse during a given hour). All analyses were stratified by patients' initial hospital unit (ward, stepdown unit, or intensive care unit). We used discrete-time hazard regression to assess the association between each novel time-varying predictor and the outcomes of discharge and mortality, separately.

RESULTS

Our dataset consisted of 84 162 161 hourly records from 954 477 hospitalizations. Many novel time-varying predictors had strong associations with the 2 study outcomes. However, most of the predictors did not merely track patients' severity of illness; instead, many of them only had weak correlations with severity, often with complex relationships over time.

DISCUSSION

Increasing availability of process of care data from automated electronic health records will permit better quantification of healthcare delivery strain. This could result in enhanced prediction of adverse outcomes and service delays.

CONCLUSION

New conceptual models will be needed to use these new data elements.

Do-not-resuscitate orders in patients with community-acquired pneumonia: a retrospective study

Background

To investigate the use of do-not-resuscitate (DNR) orders in patients hospitalized with community-acquired pneumonia (CAP) and the association with mortality.

Methods

We assembled a cohort of 1317 adults hospitalized with radiographically confirmed CAP in three Danish hospitals. Patients were grouped into no DNR order, early DNR order (≤48 h after admission), and late DNR order (> 48 h after admission). We tested for associations between a DNR order and mortality using a cox proportional hazard model adjusted for patient and disease related factors.

Results

Among 1317 patients 177 (13%) patients received a DNR order: 107 (8%) early and 70 (5%) late, during admission. Patients with a DNR order were older (82 years vs. 70 years, p < 0.001), more frequently nursing home residents (41% vs. 6%, p < 0.001) and had more comorbidities (one or more comorbidities: 73% vs. 59%, p < 0.001). The 30-day mortality was 62% and 4% in patients with and without a DNR order, respectively. DNR orders were associated with increased risk of 30-day mortality after adjustment for age, nursing home residency and comorbidities. The association was modified by the CURB-65 score Hazard ratio (HR) 39.3 (95% CI 13.9–110.6), HR 24.0 (95% CI 11.9–48,3) and HR 9.4 (95% CI: 4.7–18.6) for CURB-65 score 0–1, 2 and 3–5, respectively.

Conclusion

In this representative Danish cohort, 13% of patients hospitalized with CAP received a DNR order. DNR orders were associated with higher mortality after adjustment for clinical risk factors. Thus, we encourage researcher to take DNR orders into account as potential confounder when reporting CAP associated mortality.

Effect of a rapid response system on code rates and in-hospital mortality in medical wards

BACKGROUND

To determine the effects of implementing a rapid response system (RRS) on code rates and in-hospital mortality in medical wards.

METHODS

This retrospective study included adult patients admitted to medical wards at Seoul National University Hospital between July 12, 2016 and March 12, 2018; the sample comprised 4,224 patients admitted 10 months before RRS implementation and 4,168 patients admitted 10 months following RRS implementation. Our RRS only worked during the daytime (7 AM to 7 PM) on weekdays. We compared code rates and in-hospital mortality rates between the preintervention and postintervention groups.

RESULTS

There were 62.3 RRS activations per 1,000 admissions. The most common reasons for RRS activation were tachypnea or hypopnea (44%), hypoxia (31%), and tachycardia or bradycardia (21%). Code rates from medical wards during RRS operating times significantly decreased from 3.55 to 0.96 per 1,000 admissions (adjusted odds ratio [aOR], 0.29; 95% confidence interval [CI], 0.10 to 0.87; P=0.028) after RRS implementation. However, code rates from medical wards during RRS nonoperating times did not differ between the preintervention and postintervention groups (2.60 vs. 3.12 per 1,000 admissions; aOR, 1.23; 95% CI, 0.55 to 2.76; P=0.614). In-hospital mortality significantly decreased from 56.3 to 42.7 per 1,000 admissions after RRS implementation (aOR, 0.79; 95% CI, 0.64 to 0.97; P=0.024).

CONCLUSIONS

Implementation of an RRS was associated with significant reductions in code rates during RRS operating times and in-hospital mortality in medical wards.

Do-Not-Resuscitate Status and Risk-Standardized Mortality and Readmission Rates Following Acute Myocardial Infarction

Background Important administrative-based measures of hospital quality, including those used by Centers for Medicare and Medicaid Services, may not adequately account for patient illness and social factors that vary between hospitals and can strongly affect outcomes. Do-not-resuscitate (DNR) order on admission (within the first 24 hours) is one such factor that may reflect higher preadmission illness burden as well as patients' desire for less-intense therapeutic interventions and has been shown to vary widely between hospitals. We sought to evaluate how accounting for early DNR affected hospital quality measures for acute myocardial infarction. Methods AND RESULTS We identified all patients admitted with acute myocardial infarction using the California State Inpatient Database, which captures early DNR use within 24 hours of admission. We generated hospital risk-standardized mortality and readmissions using random-effects logistic regression, before and after including early DNR status, to examine changes in overall model fit and hospital outlier designations. We included 109 521 patients from 289 hospitals and found that 8.5% (9356) patients had early DNR. Early DNR use varied widely, with median (interquartile range) hospital rates of 7.9% (4.1%-14.0%). Including early DNR in models used to assess hospital quality resulted in improvement in the mortality model (C statistics from 0.754 [0.748-0.759] to 0.784 [0.779-0.789]) but not the readmissions model. Of the hospitals designated high outliers for mortality and readmissions by the Centers for Medicare/Medicaid Services model, and therefore destined for a financial penalty, 6/25 (24%) were reclassified as nonoutliers for mortality and 2/14 (14.3%) for readmissions after including DNR status. Agreement in outlier status between the models before and after inclusion of early DNR status was moderate for mortality (κ, 0.603 [0.482-0.724]; P<0.001) and high for readmissions (κ, 0.888 [0.800-0.977]; P<0.001). Conclusions Including early DNR status in risk-adjustment models significantly improved model fit and resulted in substantial reclassification of hospital performance rankings for mortality and moderate reclassification for readmissions. DNR status at hospital admission should be considered when reporting risk-standardized hospital mortality.

Comparing Hospital Processes and Outcomes in California Medicare Beneficiaries: Simulation Prompts Reconsideration

INTRODUCTION

This article is not a traditional research report. It describes how conducting a specific set of benchmarking analyses led us to broader reflections on hospital benchmarking. We reexamined an issue that has received far less attention from researchers than in the past: How variations in the hospital admission threshold might affect hospital rankings. Considering this threshold made us reconsider what benchmarking is and what future benchmarking studies might be like. Although we recognize that some of our assertions are speculative, they are based on our reading of the literature and previous and ongoing data analyses being conducted in our research unit. We describe the benchmarking analyses that led to these reflections.

OBJECTIVES

The Centers for Medicare and Medicaid Services' Hospital Compare Web site includes data on fee-for-service Medicare beneficiaries but does not control for severity of illness, which requires physiologic data now available in most electronic medical records.To address this limitation, we compared hospital processes and outcomes among Kaiser Permanente Northern California's (KPNC) Medicare Advantage beneficiaries and non-KPNC California Medicare beneficiaries between 2009 and 2010.

METHODS

We assigned a simulated severity of illness measure to each record and explored the effect of having the additional information on outcomes.

RESULTS

We found that if the admission severity of illness in non-KPNC hospitals increased, KPNC hospitals' mortality performance would appear worse; conversely, if admission severity at non-KPNC hospitals' decreased, KPNC hospitals' performance would appear better.

CONCLUSION

Future hospital benchmarking should consider the impact of variation in admission thresholds.

Relationship between Hospital 30-Day Mortality Rates for Heart Failure and Patterns of Early Inpatient Comfort Care

BACKGROUND

The Centers for Medicare & Medicaid Services rewards hospitals that have low 30-day riskstandardized mortality rates (RSMR) for heart failure (HF).

OBJECTIVE

To describe the use of early comfort care for patients with HF, and whether hospitals that more commonly initiate comfort care have higher 30-day mortality rates.

DESIGN

A retrospective, observational study.

SETTING

Acute care hospitals in the United States.

PATIENTS

A total of 93,920 fee-for-service Medicare beneficiaries admitted with HF from January 2008 to December 2014 to 272 hospitals participating in the Get With The Guidelines-Heart Failure registry.

EXPOSURE

Early comfort care (defined as comfort care within 48 hours of hospitalization) rate.

MEASUREMENTS

A 30-day RSMR.

RESULTS

Hospitals' early comfort care rates were low for patients admitted for HF, with no change over time (2.5% to 2.6%, from 2008 to 2014, P = .56). Rates varied widely (0% to 40%), with 14.3% of hospitals not initiating comfort care for any patients during the first 2 days of hospitalization. Risk-standardized early comfort care rates were not correlated with RSMR (median RSMR = 10.9%, 25th to 75th percentile = 10.1% to 12.0%; Spearman's rank correlation = 0.13; P = .66).

CONCLUSIONS

Hospital use of early comfort care for HF varies, has not increased over time, and on average, is not correlated with 30-day RSMR. This suggests that current efforts to lower mortality rates have not had unintended consequences for hospitals that institute early comfort care more commonly than their peers.

The Effect of Intensive Care Unit Admission Patterns on Mortality-based Critical Care Performance Measures

RATIONALE

Current mortality-based critical care performance measurement focuses on intensive care unit (ICU) admissions as a single group, conflating low-severity and high-severity ICU patients for whom performance may differ and neglecting severely ill patients treated solely on hospital wards.

OBJECTIVES

To assess the relationship between hospital performance as measured by risk-standardized mortality for severely ill ICU patients, less severely ill ICU patients, and severely ill patients outside the ICU.

METHODS

Using a statewide, all-payer dataset from the Pennsylvania Healthcare Cost Containment Council, we analyzed discharge data for patients with nine clinical conditions with frequent ICU use. Using a validated severity-of-illness measure, we categorized hospitalized patients as either high severity (predicted probability of in-hospital death in top quartile) or low severity (all others). We then created three mutually exclusive groups: high-severity ICU admissions, low-severity ICU admissions, and high-severity ward patients. We used hierarchical logistic regression to generate hospital-specific 30-day risk-standardized mortality rates for each group and then compared hospital performance across groups using Spearman's rank correlation.

MEASUREMENTS AND MAIN RESULTS

We analyzed 87 hospitals with 22,734 low-severity ICU admissions (mean per hospital, 261 ± 187), 10,991 high-severity ICU admissions (mean per hospital, 126 ± 105), and 6,636 high-severity ward patients (mean per hospital, 76 ± 48). We found little correlation between hospital performance for high-severity ICU patients versus low-severity ICU patients (ρ = 0.15; P = 0.17). There were 29 hospitals (33%) that moved up or down at least two quartiles of performance across the ICU groups. There was weak correlation between hospital performance for high-severity ICU patients versus high-severity ward patients (ρ = 0.25; P = 0.02). There were 24 hospitals (28%) that moved up or down at least two quartiles of performance across the high-severity groups.

CONCLUSIONS

Hospitals that perform well in caring for high-severity ICU patients do not necessarily also perform well in caring for low-severity ICU patients or high-severity ward patients, indicating that risk-standardized mortality rates for ICU admissions as a whole offer only a narrow window on a hospital's overall performance for critically ill patients.

Accounting for Patient Preferences Regarding Life-Sustaining Treatment in Evaluations of Medical Effectiveness and Quality

The importance of understanding patient preferences for life-sustaining treatment is well described for individual clinical decisions; however, its role in evaluations of healthcare outcomes and quality has received little attention. Decisions to limit life-sustaining therapies are strongly associated with high risks for death in ways that are unaccounted for by routine measures of illness severity. However, this essential information is generally unavailable to researchers, with the potential for spurious inferences. This may lead to "confounding by unmeasured patient preferences" (a type of confounding by indication) and has implications for assessments of treatment effectiveness and healthcare quality, especially in acute and critical care settings in which risk for death and adverse events are high. Through a collection of case studies, we explore the effect of unmeasured patient resuscitation preferences on issues critical for researchers and research consumers to understand. We then propose strategies to more consistently elicit, record, and harmonize documentation of patient preferences that can be used to attenuate confounding by unmeasured patient preferences and provide novel opportunities to improve the patient centeredness of medical care for serious illness.

Variation in Do-Not-Resuscitate Orders and Implications for Heart Failure Risk-Adjusted Hospital Mortality Metrics

OBJECTIVES

This study evaluated the effect of patient do-not-resuscitate (DNR) status on hospital risk-adjusted heart failure mortality metrics.

BACKGROUND

Do-not-resuscitate orders limit the use of life-sustaining therapies. Patients with DNR orders have increased in-hospital mortality, and DNR rates vary among hospitals. Variations in DNR rates could strongly confound risk-adjusted hospital mortality rates for heart failure.

METHODS

We identified a cohort of adults with primary diagnosis of heart failure by using the 2011 California State Inpatient Database, a claims database that captures "early DNR," within 24 h of admission. Hospital-level risk-standardized in-hospital mortality was determined using random effects logistic regression. We explored changes in outlier status in models with and without early DNR status.

RESULTS

Among 55,865 patients from 290 hospitals hospitalized with heart failure, 12.1% (11.8% to 12.4%) had an early DNR order. Hospitals with higher risk-standardized DNR rates had higher risk-standardized mortality (ρ = 0.241; 95% confidence interval [CI]: 0.129 to 0.346; p < 0.001). Including DNR in models used to benchmark hospital mortality improved model performance (c-statistic from 0.821 [95% CI: 0.812 to 0.830] to 0.845 [95% CI: 0.837 to 0.853]; increased model explanatory power by 17%). Including DNR resulted in reclassification of 9.3% of hospitals' outlier status. Agreement in hospital outlier designation between models with and without DNR was low to moderate (kappa coefficient: 0.492; 95% CI: 0.331 to 0.654).

CONCLUSIONS

Accounting for DNR status resulted in a change in estimated risk-standardized mortality rates and classification of hospitals as performance "outliers." Given public reporting of heart failure mortality measurements and their influence on reimbursement, accounting for the presence of early DNR orders in quality measures should be considered.

Early transition to comfort measures only in acute stroke patients: Analysis from the Get With The Guidelines-Stroke registry

BACKGROUND

Death after acute stroke often occurs after forgoing life-sustaining interventions. We sought to determine the patient and hospital characteristics associated with an early decision to transition to comfort measures only (CMO) after ischemic stroke (IS), intracerebral hemorrhage (ICH), and subarachnoid hemorrhage (SAH) in the Get With The Guidelines-Stroke registry.

METHODS

We identified patients with IS, ICH, or SAH between November 2009 and September 2013 who met study criteria. Early CMO was defined as the withdrawal of life-sustaining treatments and interventions by hospital day 0 or 1. Using multivariable logistic regression, we identified patient and hospital factors associated with an early (by hospital day 0 or 1) CMO order.

RESULTS

Among 963,525 patients from 1,675 hospitals, 54,794 (5.6%) had an early CMO order (IS: 3.0%; ICH: 19.4%; SAH: 13.1%). Early CMO use varied widely by hospital (range 0.6%-37.6% overall) and declined over time (from 6.1% in 2009 to 5.4% in 2013; p < 0.001). In multivariable analysis, older age, female sex, white race, Medicaid and self-pay/no insurance, arrival by ambulance, arrival off-hours, baseline nonambulatory status, and stroke type were independently associated with early CMO use (vs no early CMO). The correlation between hospital-level risk-adjusted mortality and the use of early CMO was stronger for SAH (r = 0.52) and ICH (r = 0.50) than AIS (r = 0.15) patients.

CONCLUSIONS

Early CMO was utilized in about 5% of stroke patients, being more common in ICH and SAH than IS. Early CMO use varies widely between hospitals and is influenced by patient and hospital characteristics.

Hospital Variation in Utilization of Life-Sustaining Treatments among Patients with Do Not Resuscitate Orders

OBJECTIVE

To determine between-hospital variation in interventions provided to patients with do not resuscitate (DNR) orders.

DATA SOURCES/SETTING

United States Agency of Healthcare Research and Quality, Healthcare Cost and Utilization Project, California State Inpatient Database.

STUDY DESIGN

Retrospective cohort study including hospitalized patients aged 40 and older with potential indications for invasive treatments: in-hospital cardiac arrest (indication for CPR), acute respiratory failure (mechanical ventilation), acute renal failure (hemodialysis), septic shock (central venous catheterization), and palliative care. Hierarchical logistic regression to determine associations of hospital "early" DNR rates (DNR order placed within 24 hours of admission) with utilization of invasive interventions.

DATA COLLECTION/EXTRACTION METHODS

California State Inpatient Database, year 2011.

PRINCIPAL FINDINGS

Patients with DNR orders at high-DNR-rate hospitals were less likely to receive invasive mechanical ventilation for acute respiratory failure or hemodialysis for acute renal failure, but more likely to receive palliative care than DNR patients at low-DNR-rate hospitals. Patients without DNR orders experienced similar rates of invasive interventions regardless of hospital DNR rates.

CONCLUSIONS

Hospitals vary widely in the scope of invasive or organ-supporting treatments provided to patients with DNR orders.

Association of Do-Not-Resuscitate Orders and Hospital Mortality Rate Among Patients With Pneumonia

IMPORTANCE

Hospital quality measures that do not account for patient do-not-resuscitate (DNR) status may penalize hospitals admitting a greater proportion of patients with limits on life-sustaining treatments.

OBJECTIVE

To evaluate the effect of analytic approaches accounting for DNR status on risk-adjusted hospital mortality rates and performance rankings.

DESIGN, SETTING, AND PARTICIPANTS

A retrospective, population-based cohort study was conducted among adults hospitalized with pneumonia in 303 California hospitals between January 1 and December 31, 2011. We used hierarchical logistic regression to determine associations between patient DNR status, hospital-level DNR rates, and mortality measures. Changes in hospital risk-adjusted mortality rates after accounting for patient DNR status and interhospital variation in the association between DNR status and mortality were examined. Data analysis was conducted from January 16 to September 16, 2015.

EXPOSURES

Early DNR status (within 24 hours of admission).

MAIN OUTCOMES AND MEASURES

In-hospital mortality, determined using hierarchical logistic regression.

RESULTS

A total of 90,644 pneumonia cases (5.4% of admissions) were identified among the 303 California hospitals evaluated during 2011; mean (SD) age of the patients was 72.5 (13.7) years, 51.5% were women, and 59.3% were white. Hospital DNR rates varied (median, 15.8%; 25th-75th percentile, 8.9%-22.3%). Without accounting for patient DNR status, higher hospital-level DNR rates were associated with increased patient mortality (adjusted odds ratio [OR] for highest-quartile DNR rate vs lowest quartile, 1.17; 95% CI, 1.04-1.32), corresponding to worse hospital mortality rankings. In contrast, after accounting for patient DNR status and between-hospital variation in the association between DNR status and mortality, hospitals with higher DNR rates had lower mortality (adjusted OR for highest-quartile DNR rate vs lowest quartile, 0.79; 95% CI, 0.70-0.89), with reversal of associations between hospital mortality rankings and DNR rates. Only 14 of 27 hospitals (51.9%) characterized as low-performing outliers without accounting for DNR status remained outliers after DNR adjustment. Hospital DNR rates were not significantly associated with composite quality measures of processes of care for pneumonia (r = 0.11; P = .052); however, DNR rates were positively correlated with patient satisfaction scores (r = 0.35; P < .001).

CONCLUSIONS AND RELEVANCE

Failure to account for DNR status may confound the evaluation of hospital quality using mortality outcomes, penalizing hospitals that admit a greater proportion of patients with limits on life-sustaining treatments. Stakeholders should seek to improve methods to standardize and report DNR status in hospital discharge records to allow further assessment of implications of adjusting for DNR in quality measures.

The outcome of trauma patients with do-not-resuscitate orders

BACKGROUND

Institutional variation in outcome of patients with do-not-resuscitate (DNR) orders has not been well described in the setting of trauma. The purpose of this study was to assess the impact of trauma center designation on outcome of patients with DNR orders.

MATERIALS AND METHODS

A statewide trauma database (Pennsylvania Trauma Outcome Study) was used for the analysis. Characteristics of patients with DNR orders were compared between state-designated level 1 and 2 trauma centers. Inhospital mortality and major complication rates were compared using hierarchical logistic regression models that included a random effect for trauma centers. We adjusted for a number of potential confounders and allowed for nonlinearity in injury severity score and age in these models.

RESULTS

A total of 106,291 patients (14 level 1 and 11 level 2 trauma centers) were identified in the Pennsylvania Trauma Outcome Study database between 2007 and 2011. We included 5953 patients with DNR orders (5.6%). Although more severely injured patients with comorbid disease were made DNR in level 1 trauma centers, trauma center designation level was not a significant factor for inhospital mortality of patients with DNR orders (odds ratio, 1.33; 95% confidence interval, 0.81-2.18; P = 0.26). Level 1 trauma centers were significantly associated with a higher rate of major complications (odds ratio, 1.75; 95% confidence interval, 1.11-2.75; P = 0.016).

CONCLUSIONS

Inhospital mortality of patients with DNR orders was not significantly associated with trauma designation level after adjusting for case mix. More aggressive treatment or other unknown factors may have resulted in a significantly higher complication rate at level 1 trauma centers.

"Do not resuscitate" decisions in acute respiratory distress syndrome. A secondary analysis of clinical trial data

RATIONALE

Factors and outcomes associated with end-of-life decision-making among patients during clinical trials in the intensive care unit are unclear.

OBJECTIVES

We sought to determine patterns and outcomes of Do Not Resuscitate (DNR) decisions among critically ill patients with acute respiratory distress syndrome (ARDS) enrolled in a clinical trial.

METHODS

We performed a secondary analysis of data from the ARDS Network Fluid and Catheter Treatment Trial (FACTT), collected between 2000 and 2005. We calculated mortality outcomes stratified by code status, and compared baseline characteristics of patients who became DNR during the trial with participants who remained full code.

MEASUREMENTS AND MAIN RESULTS

Among 809 FACTT participants with a code status recorded, 232 (28.7%) elected DNR status. Specifically, 37 (15.9%) chose to withhold cardiopulmonary resuscitation alone, 44 (19.0%) elected to withhold some life support measures in addition to cardiopulmonary resuscitation, and 151 (65.1%) had life support withdrawn. Admission severity of illness as measured by APACHE III score was strongly associated with election of DNR status (odds ratio, 2.2; 95% confidence interval, 1.85-2.62; P < 0.0001). Almost all (97.0%; 225 of 232) patients who selected DNR status died, and 79% (225 of 284) of patients who died during the trial were DNR. Among patients who chose DNR status but did not elect withdrawal of life support, 91% (74 of 81) died.

CONCLUSIONS

The vast majority of deaths among clinical trial patients with ARDS were preceded by a DNR order. Unlike other studies of end-of-life decision-making in the intensive care unit, nearly all patients who became DNR died. The impact of variation of practice in end-of-life decision-making during clinical trials warrants further study.

Do-not-resuscitate status and observational comparative effectiveness research in patients with septic shock*

OBJECTIVES

To assess the importance of including do-not-resuscitate status in critical care observational comparative effectiveness research.

DESIGN

Retrospective analysis.

SETTING

All California hospitals participating in the 2007 California State Inpatient Database, which provides do-not-resuscitate status within the first 24 hours of admission.

PATIENTS

Septic shock present at admission.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

We investigated the association of early do-not-resuscitate status with in-hospital mortality among patients with septic shock. We also examined the strength of confounding of do-not-resuscitate status on the association between activated protein C therapy and mortality, an association with conflicting results between observational and randomized studies. We identified 24,408 patients with septic shock; 19.6% had a do-not-resuscitate order. Compared with patients without a do-not-resuscitate order, those with a do-not-resuscitate order were significantly more likely to be older (75 ± 14 vs 67 ± 16 yr) and white (62% vs 53%), with more acute organ failures (1.44 ± 1.15 vs 1.38 ± 1.15), but fewer inpatient interventions (1.0 ± 1.0 vs 1.4 ± 1.1). Adding do-not-resuscitate status to a model with 47 covariates improved mortality discrimination (c-statistic, 0.73-0.76; p < 0.001). Addition of do-not-resuscitate status to a multivariable model assessing the association between activated protein C and mortality resulted in a 9% shift in the activated protein C effect estimate toward the null (odds ratio from 0.78 [95% CI, 0.62-0.99], p = 0.04, to 0.85 [0.67-1.08], p = 0.18).

CONCLUSIONS

Among patients with septic shock, do-not-resuscitate status acts as a strong confounder that may inform past discrepancies between observational and randomized studies of activated protein C. Inclusion of early do-not-resuscitate status into more administrative databases may improve observational comparative effectiveness methodology.

Why do nonsurvivors from community-acquired pneumonia not receive ventilatory support?

OBJECTIVE

We investigated rates and predictors of ventilatory support during hospitalization in seemingly not severely compromised nonsurvivors of community-acquired pneumonia (CAP).

METHODS

We used the database from the German nationwide mandatory quality assurance program including all hospitalized patients with CAP from 2007 to 2011. We selected a population not residing in nursing homes, not bedridden, and not referred from another hospital. Predictors of ventilatory support were identified using a multivariate analysis.

RESULTS

Overall, 563,901 patients (62.3% of the whole population) were included. Mean age was 69.4 ± 16.6 years; 329,107 (58.4%) were male. Mortality was 39,895 (7.1%). A total of 28,410 (5.0%) received ventilatory support during the hospital course, and 76.3% of nonsurvivors did not receive ventilatory support (62.6% of those aged <65 years and 78% of those aged ≥65 years). Higher age (relative risk (RR) 0.48, 95% confidence interval (CI) 0.44-0.51), failure to assess gas exchange (RR 0.18, 95% CI 0.14-0.25) and to administer antibiotics within 8 h of hospitalization (RR 0.48, 95% CI 0.39-0.59) were predictors of not receiving ventilatory support during hospitalization. Death from CAP occurred significantly earlier in the nonventilated group (8.2 ± 8.9 vs. 13.1 ± 14.1 days; p < 0.0001).

CONCLUSIONS

The number of nonsurvivors without obvious reasons for withholding ventilatory support is disturbingly high, particularly in younger patients. Both performance predictors for not being ventilated remain ambiguous, because they may reflect either treatment restrictions or deficient clinical performance. Elucidating this ambiguity will be part of the forthcoming update of the quality assurance program.

Risk-adjusting hospital mortality using a comprehensive electronic record in an integrated health care delivery system

OBJECTIVE

Using a comprehensive inpatient electronic medical record, we sought to develop a risk-adjustment methodology applicable to all hospitalized patients. Further, we assessed the impact of specific data elements on model discrimination, explanatory power, calibration, integrated discrimination improvement, net reclassification improvement, performance across different hospital units, and hospital rankings.

DESIGN

Retrospective cohort study using logistic regression with split validation.

PARTICIPANTS

A total of 248,383 patients who experienced 391,584 hospitalizations between January 1, 2008 and August 31, 2011.

SETTING

Twenty-one hospitals in an integrated health care delivery system in Northern California.

RESULTS

Inpatient and 30-day mortality rates were 3.02% and 5.09%, respectively. In the validation dataset, the greatest improvement in discrimination (increase in c statistic) occurred with the introduction of laboratory data; however, subsequent addition of vital signs and end-of-life care directive data had significant effects on integrated discrimination improvement, net reclassification improvement, and hospital rankings. Use of longitudinally captured comorbidities did not improve model performance when compared with present-on-admission coding. Our final model for inpatient mortality, which included laboratory test results, vital signs, and care directives, had a c statistic of 0.883 and a pseudo-R of 0.295. Results for inpatient and 30-day mortality were virtually identical.

CONCLUSIONS

Risk-adjustment of hospital mortality using comprehensive electronic medical records is feasible and permits one to develop statistical models that better reflect actual clinician experience. In addition, such models can be used to assess hospital performance across specific subpopulations, including patients admitted to intensive care.

Accuracy of do not resuscitate (DNR) in administrative data

This article evaluates the accuracy of reporting do not resuscitate (DNR) orders in administrative data for use in risk-adjusted hospital assessments. We compared DNR reporting by 48 California hospitals in 2005 patient discharge data (PDD) with gold-standard assessments made by registered nurses (RNs) who reabstracted 1,673 records of patients with myocardial infarction, pneumonia, or heart failure. The PDD agreed with the RN reabstraction in 1,411 (84.3%) cases. The administrative data did not reflect a DNR order in 71 of 512 records where the RN indicated there was (14% false negative rates), and reflected a DNR order in 191 of 1,161 records where the RN indicated there was not (16% false positive rate). The accuracy of DNR was more problematic for patients who died, suggesting that hospital-reported DNR is problematic for capturing patient preferences for resuscitation that can be used for risk-adjusted outcomes assessments.

Using Automated Clinical Data for Risk Adjustment

BACKGROUND

Clinically plausible risk-adjustment methods are needed to implement pay-for-performance protocols. Because billing data lacks clinical precision, may be gamed, and chart abstraction is costly, we sought to develop predictive models for mortality that maximally used automated laboratory data and intentionally minimized the use of administrative data (Laboratory Models). We also evaluated the additional value of vital signs and altered mental status (Full Models).

METHODS

Six models predicting in-hospital mortality for ischemic and hemorrhagic stroke, pneumonia, myocardial infarction, heart failure, and septicemia were derived from 194,903 admissions in 2000-2003 across 71 hospitals that imported laboratory data. Demographics, admission-based labs, International Classification of Diseases (ICD)-9 variables, vital signs, and altered mental status were sequentially entered as covariates. Models were validated using abstractions (629,490 admissions) from 195 hospitals. Finally, we constructed hierarchical models to compare hospital performance using the Laboratory Models and the Full Models.

RESULTS

Model c-statistics ranged from 0.81 to 0.89. As constructed, laboratory findings contributed more to the prediction of death compared with any other risk factor characteristic groups across most models except for stroke, where altered mental status was more important. Laboratory variables were between 2 and 67 times more important in predicting mortality than ICD-9 variables. The hospital-level risk-standardized mortality rates derived from the Laboratory Models were highly correlated with the results derived from the Full Models (average rho = 0.92).

CONCLUSIONS

Mortality can be well predicted using models that maximize reliance on objective pathophysiologic variables whereas minimizing input from billing data. Such models should be less susceptible to the vagaries of billing information and inexpensive to implement.