Associations between Present-on-Admission Do-Not-Resuscitate Orders and Short-Term Outcomes in Patients with Pneumonia.

OBJECTIVES: Do-not-resuscitate (DNR) orders are used to express patient preferences for cardiopulmonary resuscitation. This study examined whether early DNR orders are associated with differences in treatments and outcomes among patients hospitalized with pneumonia. METHODS: This is a retrospective cohort study of 768,015 adult patients hospitalized with pneumonia from 2010 to 2015 in 646 US hospitals. The exposure was DNR orders present on admission. Secondary analyses stratified patients by predicted in-hospital mortality. Main outcomes included in-hospital mortality, length of stay, cost, intensive care admission, invasive mechanical ventilation, noninvasive ventilation, vasopressors, and dialysis initiation. RESULTS: Of 768,015 patients, 94,155 (12.3%) had an early DNR order. Compared with those without, patients with DNR orders were older (mean age 80.1 ± 10.6 years vs 67.8 ± 16.4 years), with higher comorbidity burden, intensive care use (31.6% vs 30.6%), and in-hospital mortality (28.2% vs 8.5%). After adjustment via propensity score weighting, these patients had higher mortality (odds ratio [OR] 2.39, 95% confidence interval [CI] 2.33-2.45) and lower use of intensive therapies such as vasopressors (OR 0.83, 95% CI 0.81-0.85) and invasive mechanical ventilation (OR 0.68, 95% CI 0.66-0.70). Although there was little relationship between predicted mortality and DNR orders, among those with highest predicted mortality, DNR orders were associated with lower intensive care use compared with those without (66.7% vs 80.8%). CONCLUSIONS: Patients with early DNR orders have higher in-hospital mortality rates than those without, but often receive intensive care. These orders have the most impact on the care of patients with the highest mortality risk.

Derivation and validation of a risk assessment model for drug-resistant pathogens in hospitalized patients with community-acquired pneumonia.

OBJECTIVE: To derive and validate a model for risk of resistance to first-line community-acquired pneumonia (CAP) therapy. DESIGN: We developed a logistic regression prediction model from a large multihospital discharge database and validated it versus the Drug Resistance in Pneumonia (DRIP) score in a holdout sample and another hospital system outside that database. Resistance to first-line CAP therapy (quinolone or third generation cephalosporin plus macrolide) was based on blood or respiratory cultures. SETTING: This study was conducted using data from 177 Premier Healthcare database hospitals and 11 Cleveland Clinic hospitals. PARTICIPANTS: Adults hospitalized for CAP. EXPOSURE: Risk factors for resistant infection. RESULTS: Among 138,762 eligible patients in the Premier database, 12,181 (8.8%) had positive cultures and 5,200 (3.8%) had organisms resistant to CAP therapy. Infection with a resistant organism in the previous year was the strongest predictor of resistance; markers of acute illness (eg, receipt of mechanical ventilation or vasopressors) and chronic illness (eg, pressure ulcer, paralysis) were also associated with resistant infections. Our model outperformed the DRIP score with a C-statistic of 0.71 versus 0.63 for the DRIP score (P < .001) in the Premier holdout sample, and 0.65 versus 0.58 (P < .001) in Cleveland Clinic hospitals. Clinicians at Premier facilities used broad-spectrum antibiotics for 20%-30% of patients. In discriminating between patients with and without resistant infections, physician judgment slightly outperformed the DRIP instrument but not our model. CONCLUSIONS: Our model predicting infection with a resistant pathogen outperformed both the DRIP score and physician practice in an external validation set. Its integration into practice could reduce unnecessary use of broad-spectrum antibiotics.

A risk model to identify Legionella among patients admitted with community-acquired pneumonia: A retrospective cohort study.

BACKGROUND: Guidelines recommend testing hospitalized patients with community-acquired pneumonia (CAP) for Legionella pneumophila only if the infection is severe or risk factors are present. There are no validated models for predicting Legionella. OBJECTIVE: To derive and externally validate a model to predict a positive Legionella test. DESIGN, SETTING AND PARTICIPANTS: Diagnostic study of adult inpatients with pneumonia using data from 177 US hospitals in the Premier Healthcare Database (training and hold-out validation sets) and 12 Cleveland Clinic Health System (CCHS) hospitals (external validation set). We used multiple logistic regression to predict positive Legionella tests in the training set, and evaluated performance in both validation sets. MAIN OUTCOME AND MEASURES: The outcome was a positive Legionella test. Potential predictors included demographics and co-morbidities, disease severity indicators, season, region, and presence of a local outbreak. RESULTS: Of 166,689 patients hospitalized for pneumonia, 43,070 were tested for Legionella and 642 (1.5%) tested positive. The strongest predictors of a positive test were a local outbreak (odds ratio [OR], 3.4), June-October occurrence (OR, 3.4), hyponatremia (OR, 3.3), smoking (OR, 2.4), and diarrhea (OR, 2.0); prior admission within 6 months (OR, 0.27) and chronic pulmonary disease (OR, 0.49) were associated with a negative test. Model c-statistics were 0.79 in the Premier and 0.77 in the CCHS validation samples. High-risk patients were only slightly more likely to have been tested than lower-risk patients. Compared to actual practice, the model-based testing strategy detected twice as many cases. CONCLUSIONS: Although Legionella is an uncommon cause of pneumonia, patient characteristics can identify individuals at high risk, allowing for more efficient testing.

Benchmarking Inpatient Mortality Using Electronic Medical Record Data: A Retrospective, Multicenter Analytical Observational Study.

OBJECTIVES: To develop a model to benchmark mortality in hospitalized patients using accessible electronic medical record data. DESIGN: Univariate analysis and multivariable logistic regression were used to identify variables collected during the first 24 hours following admission to test for risk factors associated with the end point of hospital mortality. Models were built using specific diagnosis (International Classification of Diseases, 9th Edition or International Classification of Diseases, 10th Edition) captured at discharge, rather than admission diagnosis, which may be discordant. Variables were selected based, in part, on prior the Acute Physiology and Chronic Health Evaluation methodology and included primary diagnosis information plus three aggregated indices (physiology, comorbidity, and support). A Physiology Index was created using parsimonious nonlinear modeling of heart rate, mean arterial pressure, temperature, respiratory rate, hematocrit, platelet counts, and serum sodium. A Comorbidity Index incorporates new or ongoing diagnoses captured by the electronic medical record during the preceding year. A Support Index considered 10 interventions such as mechanical ventilation, selected IV drugs, and hemodialysis. Accuracy was determined using area under the receiver operating curve for discrimination, calibration curves, and modified Brier score for calibration. SETTING AND PATIENTS: We used deidentified electronic medical record data from 74,434 adult inpatients (ICU and ward) at 15 hospitals from 2010 to 2013 to develop the mortality model and validated using data for additional 49,752 patients from the same 15 hospitals. A second revalidation was accomplished using data on 83,684 patients receiving care at six hospitals between 2014 and 2016. The model was also validated on a subset of patients with an ICU stay on day 1. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: This model uses physiology, comorbidity, and support indices, primary diagnosis, age, lowest Glasgow Coma Score, and elapsed time since hospital admission to predict hospital mortality. In the initial validation cohort, observed mortality was 4.04% versus predicted mortality 4.12% (Student t test, p = 0.37). In the revalidation using a different set of hospitals, predicted and observed mortality were 2.66% and 2.99%, respectively. Area under the receiver operating curve were 0.902 (0.895-0.909) and 0.884 (0.877-0.891), respectively, and calibration curves show a close relationship of observed and predicted mortalities. In the evaluation of the subset of ICU patients on day1, the area under the receiver operating curve was 0.87, with an observed mortality of 8.78% versus predicted mortality of 8.93% (Student t test, p = 0.52) and a standardized mortality ratio of 0.98 (0.932-1.034). CONCLUSIONS: Variables considered by traditional ICU prognostic models accurately benchmark patient mortality for patients receiving care in multiple hospital locations, not only the ICU. Unlike Acute Physiology and Chronic Health Evaluation, this model relies on electronic medical record data alone and does not require personnel to collect the independent predictor variables. Assessing the model's utility for benchmarking hospital performance will require prospective testing in a larger representative sample of hospitals.

Coronavirus Disease 2019 ICU Patients Have Higher-Than-Expected Acute Physiology and Chronic Health Evaluation-Adjusted Mortality and Length of Stay Than Viral Pneumonia ICU Patients.

OBJECTIVES: To compare Acute Physiology and Chronic Health Evaluation-IV-adjusted mortality and length of stay outcomes of adult ICU patients who tested positive for coronavirus disease 2019 with patients admitted to ICU with other viral pneumonias including a subgroup with viral pneumonia and concurrent acute respiratory distress syndrome (viral pneumonia-acute respiratory distress syndrome). DESIGN: Retrospective review of Acute Physiology and Chronic Health Evaluation data collected from routine clinical care. SETTING: Forty-three hospitals contributing coronavirus disease 2019 patient data between March 14, and June 17, 2020, and 132 hospitals in the United States contributing data on viral pneumonia patients to the Acute Physiology and Chronic Health Evaluation database between January 1, 2014, and December 31, 2019. PATIENTS AND MEASUREMENTS: One thousand four hundred ninety-one patients with diagnosis of coronavirus disease 2019 infection and 4,200 patients with a primary (n = 2,544) or secondary (n = 1,656) admitting diagnosis of noncoronavirus disease viral pneumonia receiving ICU care. A subset of 202 viral pneumonia patients with concurrent acute respiratory distress syndrome was examined separately. INTERVENTIONS: None. MAIN RESULTS: Mean age was 63.4 for coronavirus disease (p = 0.064) versus 64.1 for viral pneumonia. Acute Physiology and Chronic Health Evaluation-IV scores were similar at 56.7 and 55.0, respectively (p = 0.060), but gender and ethnic distributions differed, as did Pao2 to Fio2 ratio and WBC count at admission. The hospital standardized mortality ratio (95% CI) was 1.52 (1.35-1.68) for coronavirus disease patients and 0.82 (0.75-0.90) for viral pneumonia patients. In the coronavirus disease group, ICU and hospital length of stay were 3.1 and 3.0 days longer than in viral pneumonia patients. Standardized ICU and hospital length of stay ratios were 1.13 and 1.46 in the coronavirus disease group versus 0.95 and 0.94 in viral pneumonia patients. Forty-seven percent of coronavirus disease patients received invasive or noninvasive ventilatory support on their first ICU day versus 65% with viral pneumonia. Ventilator days in survivors were longer in coronavirus disease (10.4 d) than in viral pneumonia (4.3 d) patients, except in the viral pneumonia-acute respiratory distress syndrome subgroup (10.2 d). CONCLUSIONS: Severity-adjusted mortality and length of stay are higher for coronavirus disease 2019 patients than for viral pneumonia patients admitted to ICU. Coronavirus disease patients also have longer time on ventilator and ICU length of stay, comparable with the subset of viral pneumonia patients with concurrent acute respiratory distress syndrome. Mortality and length of stay increase with age and higher scores in both populations, but observed to predicted mortality and length of stay are higher than expected with coronavirus disease patients across all severity of illness levels. These findings have implications for benchmarking ICU outcomes during the coronavirus disease 2019 pandemic.

Assessment of the Accuracy of Using ICD-9 Diagnosis Codes to Identify Pneumonia Etiology in Patients Hospitalized With Pneumonia.

Importance: Administrative databases may offer efficient clinical data collection for studying epidemiology, outcomes, and temporal trends in health care delivery. However, such data have seldom been validated against microbiological laboratory results. Objective: To assess the validity of International Classification of Diseases, Ninth Revision (ICD-9) organism-specific administrative codes for pneumonia using microbiological data (test results for blood or respiratory culture, urinary antigen, or polymerase chain reaction) as the criterion standard. Design, Setting, and Participants: Cross-sectional diagnostic accuracy study conducted between February 2017 and June 2019 using data from 178 US hospitals in the Premier Healthcare Database. Patients were aged 18 years or older admitted with pneumonia and discharged between July 1, 2010, and June 30, 2015. Data were analyzed from February 14, 2017, to June 27, 2019. Exposures: Organism-specific pneumonia identified from ICD-9 codes. Main Outcomes and Measures: Sensitivity, specificity, positive predictive value, and negative predictive value of ICD-9 codes using microbiological data as the criterion standard. Results: Of 161 529 patients meeting inclusion criteria (mean [SD] age, 69.5 [16.2] years; 51.2% women), 35 759 (22.1%) had an identified pathogen. ICD-9-coded organisms and laboratory findings differed notably: for example, ICD-9 codes identified only 14.2% and 17.3% of patients with laboratory-detected methicillin-sensitive Staphylococcus aureus and Escherichia coli, respectively. Although specificities and negative predictive values exceeded 95% for all codes, sensitivities ranged downward from 95.9% (95% CI, 95.3%-96.5%) for influenza virus to 14.0% (95% CI, 8.8%-20.8%) for parainfluenza virus, and positive predictive values ranged downward from 91.1% (95% CI, 89.5%-92.6%) for Staphylococcus aureus to 57.1% (95% CI, 39.4%-73.7%) for parainfluenza virus. Conclusions and Relevance: In this study, ICD-9 codes did not reliably capture pneumonia etiology identified by laboratory testing; because of the high specificities of ICD-9 codes, however, administrative data may be useful in identifying risk factors for resistant organisms. The low sensitivities of the diagnosis codes may limit the validity of organism-specific pneumonia prevalence estimates derived from administrative data.

Pressure Injuries at Intensive Care Unit Admission as a Prognostic Indicator of Patient Outcomes.

BACKGROUND: Pressure injuries, also known as pressure ulcers, are a serious complication of immobility. Patients should be thoroughly examined for pressure injuries when admitted to the intensive care unit to optimize treatment. Whether community-acquired pressure injuries correlate with poor hospital outcomes among critically ill patients is understudied. OBJECTIVES: To determine whether pressure injuries present upon admission to the intensive care unit can serve as a predictive marker for longer hospitalization and increased mortality. METHODS: This study retrospectively analyzed admissions of adult patients to a 24-bed medical-surgical intensive care unit in a large level I trauma center in the northeast United States from 2010 to 2012. The association of pressure injuries with mortality and length of stay was assessed, using multivariable logistic regression and generalized linear models, adjusted for age, comorbidities, Acute Physiology and Chronic Health Evaluation III score, and other patient characteristics. RESULTS: Among 2723 patients, 180 (6.6%) had a pressure injury at admission. Patients with a pressure injury had longer mean unadjusted stay (15.6 vs 10.5 days; P < .001) and higher in-hospital mortality rate (32.2% vs 18.3%; P < .001) than did patients without a pressure injury at admission. After multivariable adjustment, pressure injuries were associated with a mean increase in length of stay of 3.1 days (95% CI 1.5-4.7; P < .001). Pressure injuries were not associated with mortality after adjusting for the Acute Physiology and Chronic Health Evaluation III score, but they may serve as a marker for increased risk of mortality if an Acute Physiology and Chronic Health Evaluation III score is unavailable. CONCLUSION: Pressure injuries present at admission to the intensive care unit are an objective, easy-to-identify finding associated with longer stays. Pressure injuries might have a modest association with higher risk of mortality.

Association Between Alcohol Use Disorders and Outcomes of Patients Hospitalized With Community-Acquired Pneumonia.

Importance: Patients with alcohol use disorder (AUD) are at elevated risk of developing pneumonia, but few studies have assessed the outcomes of pneumonia in patients with AUD. Objectives: To compare the causes, treatment, and outcomes of pneumonia in patients with and without AUD and to understand the associations of comorbid illnesses, alcohol withdrawal, and any residual effects due to alcohol itself with patient outcomes. Design, Setting, and Participants: A retrospective cohort study was conducted of 137 496 patients 18 years or older with pneumonia who were admitted to 177 US hospitals participating in the Premier Healthcare Database from July 1, 2010, to June 30, 2015. Statistical analysis was conducted from October 27, 2017, to August 20, 2018. Exposure: Alcohol use disorders identified from International Classification of Diseases, Ninth Revision, Clinical Modification codes. Main Outcomes and Measures: Pneumonia cause, antibiotic treatment, inpatient mortality, clinical deterioration, length of stay, and cost. Associations of AUD with these variables were studied using generalized linear mixed models. Results: Of 137 496 patients with community-acquired pneumonia (70 358 women and 67 138 men; mean [SD] age, 69.5 [16.2] years), 3.5% had an AUD. Patients with an AUD were younger than those without an AUD (median age, 58.0 vs 73.0 years; P < .001), more often male (77.3% vs 47.8%; P < .001), and more often had principal diagnoses of aspiration pneumonia (10.9% vs 9.8%; P < .001), sepsis (38.6% vs 30.7%; P < .001), or respiratory failure (9.3% vs 5.5%; P < .001). Their cultures more often grew Streptococcus pneumoniae (43.7% vs 25.5%; P < .001) and less frequently grew organisms resistant to guideline-recommended antibiotics (25.0% vs 43.7%; P < .001). Patients with an AUD were treated more often with piperacillin-tazobactam (26.2% vs 22.5%; P < .001) but equally as often with anti-methicillin-resistant Staphylococcus aureus agents (32.9% vs 31.8%; P = .11) compared with patients without AUDs. When adjusted for demographic characteristics and insurance, AUD was associated with higher mortality (odds ratio, 1.40; 95% CI, 1.25-1.56), length of stay (risk-adjusted geometric mean ratio, 1.24; 95% CI, 1.20-1.27), and costs (risk-adjusted geometric mean ratio, 1.33; 95% CI, 1.28-1.38). After additional adjustment for differences in comorbidities and risk factors for resistant organisms, AUD was no longer associated with mortality but remained associated with late mechanical ventilation (odds ratio, 1.28; 95% CI, 1.12-1.46), length of stay (risk-adjusted geometric mean ratio, 1.04; 95% CI, 1.01-1.06), and costs (risk-adjusted geometric mean ratio, 1.06; 95% CI, 1.03-1.09). Models segregating patients undergoing alcohol withdrawal showed that poorer outcomes among patients with AUD were confined to the subgroup undergoing alcohol withdrawal. Conclusions and Relevance: This study suggests that, compared with hospitalized patients with community-acquired pneumonia but without AUD, those with AUD less often harbor resistant organisms. The higher age-adjusted risk of death among patients with AUD appears to be largely attributable to differences in comorbidities, whereas greater use of health care resources may be attributable to alcohol withdrawal.

An analysis of homeless patients in the United States requiring ICU admission.

PURPOSE: To assess how homelessness impacts mortality and length of stay (LOS) among select the intensive care unit (ICU) patients. METHODS: We used ICD-9 code V60.0 to identify homeless patients using the Premier Perspective Database from January 2010 to June 2011. We identified three subpopulations who received critical care services using ICD-9 and Medicare Severity Diagnosis Related Groups (MS-DRG) codes: patients with a diagnosis of sepsis who were treated with antibiotics by Day 2, patients with an alcohol or drug related MS-DRG, and patients with a diabetes related MS-DRG. We used multivariable logistic regression to predict mortality and multivariable generalized estimating equations to predict hospital and ICU LOS. RESULTS: 781,540 hospitalizations met inclusion criteria; 2278 (0.3%) were homeless. We found homelessness had no significant adjusted association with mortality among sepsis patients, but was associated with substantially longer hospital LOS: (3.7 days longer; 95% CI (1.7, 5.7, p < .001). LOS did not differ in the Diabetes or Alcohol and Drug related DRG groups. CONCLUSIONS: Critically ill homeless patients with sepsis had longer hospital LOS but similar ICU LOS and mortality risk compared to non-homeless patients. Homelessness was not associated with increased LOS in the diabetes or alcohol and drug related groups.

Treatment Trends and Outcomes in Healthcare-Associated Pneumonia.

BACKGROUND: The American Thoracic Society and Infectious Diseases Society of America guidelines for management of healthcare-associated pneumonia (HCAP), first published in 2005, have been controversial regarding the selection of empiric broad-spectrum antibiotics, whether the criteria for HCAP predicts the likelihood of infection with multidrug resistant organisms, and whether HCAP patients have improved outcomes when treated with empiric broad-spectrum antibiotics. METHODS: A retrospective cohort study at 488 US hospitals from July 2007 to November 2011. Patients who met criteria for HCAP were included. Guideline-concordant antibiotics were assessed based on guideline recommendations. We assessed changes in hospital rates of concordant antibiotic use over time and their correlation with outcomes. RESULTS: Among 149,963 patients with HCAP, 19.6% received fully guideline-concordant antibiotics, 21.7% received partially concordant antibiotics, and 58.9% received discordant antibiotics. Guideline concordance increased over time. Rates of fully or partially concordant antibiotics varied across hospitals (median 36.4%; interquartile range 25.8%-49.1%). Among patients who received discordant antibiotics, 81.5% were treated according to community-acquired pneumonia (CAP) guidelines. On average, the rate of guideline concordance increased by 2.2% per 6-month interval, while hospital level rates of mortality, excess length of stay, and progression to respiratory failure did not change. CONCLUSIONS: In this large, nationally representative cohort, only 1 in 5 patients with risk factors for HCAP received treatment that was fully in accordance with guidelines, and many received CAP therapy instead. At the hospital level, increases in the use of concordant antibiotics were not associated with declines in mortality, excess length of stay, or progression to respiratory failure.

The dangers of extreme body mass index values in patients with Clostridium difficile.

PURPOSE: To examine the association between body mass index (BMI) and in-hospital mortality in patients presenting with Clostridium difficile infections in emergency department visits (ED) in the USA. Infected patients with extreme BMIs may have an elevated mortality risk, but prior studies examining this question have been too small to reach definitive conclusions. METHODS: Data were from the Nationwide Emergency Department Sample (NEDS), Healthcare Cost and Utilization Project (HCUP), Agency for Healthcare Research and Quality during 2012. NEDS records emergency department (ED) visits across the USA and provides statistical sampling weights to approximate a nationally representative sample of US hospital-based EDs. Inclusion criteria were adults age 18 or older with an ICD-9 code for C. difficile infection (008.45) and a documented body mass index ICD-9 V code (V85.x). Logistic regression was used to predict mortality after adjusting for demographic variables and chronic comorbidities defined by Elixhauser. RESULTS: A weighted sample of 22,937 ED visits met all inclusion criteria. The cohort's mean age was 66. 64.6% were female. The unadjusted mortality rate was 6.5%. Patients with a BMI < 19 kg/m2 had an adjusted odds ratio of 2.73; 95% CI (1.80, 4.16), p < 0.001 compared to patients with a BMI of 19.0-4.9 kg/m2 (the referent category). In obese patients, only BMI values >40 kg/m2 were associated with significantly greater mortality risk. CONCLUSION: Being underweight (BMI < 19) or morbidly obese (BMI > 40) was associated with increased risk of in-hospital mortality in patients presenting with C. difficile infections.

Association Between Initial Route of Fluoroquinolone Administration and Outcomes in Patients Hospitalized for Community-acquired Pneumonia.

BACKGROUND: Fluoroquinolones have equivalent oral and intravenous bioavailability, but hospitalized patients with community-acquired pneumonia (CAP) generally are treated intravenously. Our objectives were to compare outcomes of hospitalized CAP patients initially receiving intravenous vs oral respiratory fluoroquinolones. METHODS: This was a retrospective cohort study utilizing data from 340 hospitals involving CAP patients admitted to a non-intensive care unit (ICU) setting from 2007 to 2010, who received intravenous or oral levofloxacin or moxifloxacin. The primary outcome was in-hospital mortality. Secondary outcomes included clinical deterioration (transfer to ICU, initiation of vasopressors, or invasive mechanical ventilation [IMV] initiated after the second hospital day), antibiotic escalation, length of stay (LOS), and cost. RESULTS: Of 36 405 patients who met inclusion criteria, 34 200 (94%) initially received intravenous treatment and 2205 (6%) received oral treatment. Patients who received oral fluoroquinolones had lower unadjusted mortality (1.4% vs 2.5%; P = .002), and shorter mean LOS (5.0 vs 5.3; P < .001). Multivariable models using stabilized inverse propensity treatment weighting revealed lower rates of antibiotic escalation for oral vs intravenous therapy (odds ratio [OR], 0.84; 95% confidence interval [CI], .74-.96) but no differences in hospital mortality (OR, 0.82; 95% CI, .58-1.15), LOS (difference in days 0.03; 95% CI, -.09-.15), cost (difference in $-7.7; 95% CI, -197.4-182.0), late ICU admission (OR, 1.04; 95% CI, .80-1.36), late IMV (OR, 1.17; 95% CI, .87-1.56), or late vasopressor use (OR, 0.94; 95% CI, .68-1.30). CONCLUSIONS: Among hospitalized patients who received fluoroquinolones for CAP, there was no association between initial route of administration and outcomes. More patients may be treated orally without worsening outcomes.

A State-Level Assessment of Hospital-Based Palliative Care and the Use of Life-Sustaining Therapies in the United States.

BACKGROUND: It is unknown how the prevalence of hospitals with palliative care programs (PCPs) at the state level in the United States correlates with the treatment of critically ill patients. OBJECTIVE: We examined the relationship between state-level PCP prevalence and commonly used treatments for critically ill patients as well as other public health metrics. METHODS: We compiled state-level data for the year 2011 from multiple published sources. These included the poverty rate from the U.S. Census, public health measures such as the number of primary care physicians per 100,000 persons from America's Health Ranking website, and state-level rates for a series of validated ICD-9 (International Classification of Diseases, 9th Revision) procedure codes used for critically ill patients (e.g., prolonged acute mechanical ventilation [PAMV]) from the State Inpatient Databases (SID), Healthcare Cost and Utilization Project (HCUP), and Agency for Healthcare Research and Quality. State-level percentages of PCPs came from a published report by the Center to Advance Palliative Care (CAPC). We used the Kruskal-Wallis test and Pearson's correlation coefficient for statistical inference. RESULTS: State-level poverty rates were negatively correlated with the percent of hospitals with PCPs: r = -0.39, p = 0.005. States with more hospital-based PCPs had significantly lower rates of PAMV, tracheostomies, and hemodialysis but higher rates of nutritional support than states with fewer PCPs. CONCLUSIONS: States with more poverty and/or at high risk for delivering inefficient health care had fewer hospital PCPs. Hospital-based PCPs may influence the frequency of some interventions for critically ill patients.

Using patient admission characteristics alone to predict mortality of critically ill patients: A comparison of 3 prognostic scores.

PURPOSE: This study compared the performance of 3 admission prognostic scores in predicting hospital mortality. MATERIALS AND METHODS: Patient admission characteristics and hospital outcome of 9549 patients were recorded prospectively. The discrimination and calibration of the predicted risks of death derived from the Simplified Acute Physiology Score (SAPS III), Admission Mortality Prediction Model (MPM0 III), and admission Acute Physiology and Chronic Health Evaluation (APACHE) II were assessed by the area under the receiver operating characteristic curve and a calibration plot, respectively. MEASUREMENTS AND MAIN RESULTS: Of the 9549 patients included in the study, 1276 patients (13.3%) died after intensive care unit admission. Patient admission characteristics were significantly different between the survivors and nonsurvivors. All 3 prognostic scores had a reasonable ability to discriminate between the survivors and nonsurvivors (area under the receiver operating characteristic curve for SAPS III, 0.836; MPM0 III, 0.807; admission APACHE, 0.845), with best discrimination in emergency admissions. The SAPS III model had a slightly better calibration and overall performance (slope of calibration curve, 1.03; Brier score, 0.09; Nagelkerke R(2), 0.297) compared to the MPM0 III and admission APACHE II model. CONCLUSIONS: All 3 intensive care unit admission prognostic scores had a good ability to predict hospital mortality of critically ill patients, with best discrimination in emergency admissions.