In the News

ICU transfer, daily sedation interruption, and more.


Patients admitted from ED with certain conditions have higher risk of ICU transfer

Emergency department (ED) patients who are admitted with myocardial infarction, sepsis, or respiratory conditions have a higher risk of unplanned transfer to the intensive care unit (ICU) within 24 hours than other patients admitted from the ED, a study found.

Researchers evaluated 178,315 ED admissions at 13 Kaiser Permanente Northern California hospitals between 2007 and 2009. The main outcome was unplanned transfer to the intensive care unit (ICU) within 24 hours of admission to a hospital ward from the ED. Comparison patients were admitted to a ward from the ED but never got transferred to the ICU. Researchers examined patient and disease characteristics, hospital volume and outcomes. Results were published in the January 2013 Journal of Hospital Medicine.

Almost 2.5% (n=4,252) of evaluated patients were admitted to the ward from the ED and transferred to the ICU within 24 hours. The admitting diagnoses most associated with unplanned transfer were pneumonia (odds ratio [OR], 1.5; 95% CI, 1.2 to 1.9), myocardial infarction (MI) (OR, 1.5; 95% CI, 1.2 to 2.0), chronic obstructive pulmonary disease (COPD) (OR, 1.4; 95% CI, 1.1 to 1.9); sepsis (OR, 2.5; 95% CI, 1.9 to 3.3); and catastrophic conditions such as aortic catastrophes and shock other than septic shock (OR, 2.3; 95% CI, 1.7 to 3.0). Respiratory conditions comprised half of all diagnoses that were associated with higher risk of unplanned transfer to the ICU.

Other significant predictors of ICU transfer were being male, having a high comorbidity burden, and arriving on the ward between 11 p.m. and 7 a.m. Lower risk of ICU transfer was seen in patients admitted to monitored transitional care units (OR, 0.83; 95% CI, 0.77 to 0.90) and to higher-volume hospitals (OR, 0.94 per 1,000 additional annual ED inpatient admissions; 95% CI, 0.91 to 0.98).

The high risk of unplanned transfer for patients with respiratory conditions “indicate[s] there is some room for improvement in early care of respiratory conditions,” especially since these patients have been shown to have worse mortality than respiratory patients admitted directly to the ICU, the authors noted. Meanwhile, patients with MI, stroke and sepsis who experience unplanned ICU transfers don't have worse mortality than those with the conditions who are directly admitted. As such, closer monitoring, earlier intervention and better triage efforts to reduce unplanned transfer “would be most effective if targeted toward respiratory conditions like pneumonia and COPD,” they wrote.

Daily sedation interruption didn't benefit ventilated patients on sedation protocol

Mechanically ventilated patients receiving protocolized sedation did not benefit from daily sedation interruption in terms of intensive care unit (ICU) stay or time on ventilation, a study found.

Researchers conducted a randomized controlled trial of 430 mechanically ventilated adults from 16 tertiary care medical and surgical ICUs in the U.S. and Canada between January 2008 and July 2011. Both control and intervention groups received continuous opioid and/or benzodiazepine infusions and protocolized sedation, but the intervention group also had daily sedation interruption. Using validated scales, nurses titrated infusions to achieve light sedation. For patients receiving daily interruption, nurses resumed infusions, if indicated, at half of previous doses. The main outcome measure was time to successful extubation, with secondary outcomes including length of ICU and hospital stay, sedative and analgesic doses, delirium, unintentional device removal, and nurse and respiratory therapist clinical workload.

In both groups, the mean time to successful extubation was seven days. Length of ICU and hospital stay didn't differ significantly between the daily interruption and control groups, nor did unintentional endotracheal tube removal or delirium rates. Daily interruption was, however, associated with higher mean daily doses of midazolam (102 mg/d vs. 82 mg/d; P=0.04) and fentanyl (1,780 µg/d vs. 1,070 µg/d; P<0.001) and more daily boluses of benzodiazepines (mean, 0.253 vs. 0.177; P=0.007) and opiates (mean, 2.18 vs. 1.79; P<0.001). Self-assessed nurse workload was also higher in the interruption group (4.22 on 10-point visual-analog scale vs. 3.80; P=0.001). Results were published in the Nov. 21, 2012, Journal of the American Medical Association.

These results differ from those of two earlier trials that linked daily sedative interruption with shorter length of stay and ventilation time. The current study is different in that the protocolized sedation employed light sedation “which is likely superior to ‘usual care’ of an earlier era,” the authors noted. Also, sedation was managed by bedside ICU staff according to well-tested protocols in this study, and not by research staff as in previous studies. This trial's “unexpected” finding of greater opioid and benzodiazepine doses, and higher nurse workload, in the intervention group may be due to clinical staff's concerns about patient discomfort and stress from interrupted sedation, the authors wrote.

Majority of patients with newly diagnosed metastatic cancer may think chemotherapy curative

Large majorities of patients with metastatic lung and colorectal cancer undergoing chemotherapy believe the treatment might be curative, a study found.

Researchers applied data from the national Cancer Care Outcomes Research and Surveillance study to find out whether and why patients with metastatic disease thought chemotherapy might be curative in intent. Professional interviewers surveyed 1,193 patients four to seven months after they were diagnosed with stage IV lung or colorectal cancer and had chemotherapy. Results appeared in the Oct. 25, 2012, New England Journal of Medicine.

Overall, 69% of patients with lung cancer and 81% of those with colorectal cancer who completed the survey responded that they thought chemotherapy might be curative.

Patients in an integrated network were less likely to provide inaccurate responses (odds ratio [OR], 0.70; 95% CI, 0.52 to 0.94; P=0.02) than patients in other settings.

Patients who reported higher scores for favorable physician communication were less likely to provide accurate responses. A score of 80 to 99 resulted in less likelihood of an accurate response compared to a score of less than 80 (OR, 1.37; 95% CI, 0.93 to 2.02) and a perfect communication score of 100 resulted in less likelihood compared to a score of less than 80 (OR, 1.90; 95% CI, 1.33 to 2.72; P=0.002 for the overall comparison). Education, functional status and the patient's role in decision making were not significantly associated with the belief in curative potential.

“The observed association between inaccurate beliefs about the likelihood of cure and higher ratings of physician communication suggests a link between physicians' communication behaviors and patients' understanding of treatment benefits,” researchers wrote. “This suggests that patients perceive physicians as better communicators when they convey a more optimistic view of chemotherapy. Similarly, the finding that patients, especially those with colorectal cancer, who were treated in integrated networks were somewhat more likely to understand that chemotherapy is not curative suggests that providers may be able to improve patients' understanding if they feel it is part of their professional role.”

An editorial commented that patient misperception is likely a result of several factors, from doctors having trouble delivering news of a terminal diagnosis to patients not believing it.

“This is not one hard conversation for which we can muster our courage but a series of conversations over time from the first existential threat to life,” the editorial stated. “We recommend stating the prognosis at the first visit, appointing someone in the office to ensure there is a discussion of advance directives, helping to schedule a hospice-information visit within the first three visits, and offering to discuss prognosis and coping (‘What is important to you?’) at each transition.”

Cardiopulmonary exercise testing appears safe in high-risk patients

Cardiopulmonary exercise testing appears to be safe in patients with high-risk cardiovascular diseases, according to a study.

Researchers performed a single-center retrospective review of cardiopulmonary exercise testing in a heterogeneous cohort of high-risk patients to determine its safety in this population. A total of 5,060 exercise tests in 4,250 unique patients were included. The primary end point of the study was occurrence of major adverse events during exercise testing. Results were published in the Nov. 20, 2012, Circulation.

Of the 4,250 patients, 1,748 (35%) were women and 686 (14%) were age 75 or older, with a mean age of 57.0 ± 15.9 years. High-risk cardiac diseases included congestive heart failure (n=1,289, 25.5%), hypertropic cardiomyopathy (n=598, 11.8%), pulmonary hypertension (n=194, 3.8%) and aortic stenosis (n=212, 4.2%).

Symptom-limited exercise testing was completed in most patients (94.5%), who stopped because of fatigue, dyspnea or chest pain. Testing was stopped in the remaining 5.5% because of patient request, electrocardiogram changes, abnormal blood pressure response or major adverse event. A total of 1,192 patients (24%) were found to have a peak VO2 below 14 mL/kg/min, indicating severe functional impairment. The rate of adverse events during cardiopulmonary testing was 0.16%, most commonly ventricular tachycardia (six of eight events). Among the 10% of patients who had an implantable cardiac defibrillator, there were no discharges. No patients died during testing.

The authors noted that although many of the disorders present in their study population are usually considered contraindications to exercise testing, adverse event rates were low, and none occurred in patients with hypertrophic cardiomyopathy, pulmonary hypertension or aortic stenosis. They also noted that a panel from the Institute of Medicine has recommended cardiopulmonary exercise testing as the procedure of choice for determining cardiovascular disability, but it is usually not done in high-risk patients because of safety concerns. Their study, they said, should help change that practice.

The results have limited generalizability because the study was performed at only one center, and data on disease severity were not available for all study participants, among other limitations, the authors wrote. They also stressed that all of the patients in their study had established cardiovascular diagnoses, that tests were carefully performed at an experienced tertiary care center, and that some patients with severe forms of cardiac disease (such as severe, symptomatic aortic stenosis) should never undergo exercise testing. However, they wrote, “for other patients with these disorders, [cardiopulmonary exercise testing] appears to be reasonably safe and can serve as a helpful aid in the management of these patients.” Further study in a larger population is needed to determine which patients have the greatest risk of an adverse event during testing, they concluded.

Procalcitonin may help diagnose pneumonia when radiographs aren't definitive

Serum procalcitonin (ProCT) may help diagnose pneumonia in patients whose chest radiographs are indeterminate, a study found.

Researchers studied 528 adults admitted with acute respiratory illness to a hospital in Rochester, New York, during the winters of 2008-2009 and 2009-2010. Blood was collected in the emergency department before antibiotics were given, in order to measure ProCT. Demographic, clinical and laboratory data were collected, and the same clinical pulmonary specialist assigned a primary and secondary admission diagnosis for each patient. These diagnoses were used for the analysis but were not available to treating physicians. The serum ProCT, admission diagnoses and chest radiographic findings were used to derive receiver operating characteristic curves in order to determine the predictive accuracy of ProCT for infiltrates. Results were published in the Oct. 19, 2012, Journal of Hospital Medicine.

The top primary admission diagnoses were pneumonia (31%) and acute exacerbation of chronic obstructive pulmonary disease (AECOPD) (27%). Of 163 illnesses assigned a primary or secondary diagnosis of pneumonia, infiltrates were identified on the admission radiograph in 96% (n=156). Median ProCT levels were significantly lower in subjects with radiographs classified as “no acute disease” compared to those showing infiltrates (P<0.0001). Patients with pneumonia had higher ProCT (median 0.27 ng/mL) than those with AECOPD (0.08 ng/mL), acute bronchitis (0.09 ng/mL), or asthma (0.06 ng/mL). ProCT had moderate accuracy for the presence of infiltrates, with an area under curve of 0.72, when indeterminate radiographs were independently classified as infiltrates by a pulmonologist.

Since clinical diagnosis of pneumonia can be difficult due to ambiguous clinical and radiographic features, U.S. clinicians usually use broad-spectrum antibiotics for “possible” pneumonia, which results in overuse, the authors noted. To stem such overuse in Europe, they noted, investigators there have used ProCT successfully to guide therapy for respiratory tract infection, including pneumonia. In the U.S., “a low ProCT may help clinicians feel more confident to delay or stop antibiotics in a patient in whom clinical suspicion for bacterial pneumonia is low, yet the radiographic report states ‘pneumonia cannot be ruled out’,” the authors wrote. This could help lower unnecessary antibiotic use, they added.

Rapid and point-of-care tests accurate for HCV

Rapid diagnostic tests and point-of-care tests for hepatitis C virus (HCV) are accurate enough to be used for first-line screening, according to a recent study.

Conventional laboratory tests for HCV infection are costly, and the wait for results can be long. Researchers performed a systematic review and meta-analysis to determine whether rapid diagnostic tests and point-of-care tests, which could save time and money, have acceptable accuracy. Studies that evaluated the diagnostic accuracy of rapid and point-of-care tests for detecting HCV in patients 18 years of age and older were included in the analysis. The results appeared in the Oct. 16, 2012, Annals of Internal Medicine.

Nineteen studies were reviewed, and 18 were included in the meta-analysis. Studies were stratified by specimen type (oral fluid, whole blood, serum or plasma) or test type (rapid or point of care). High sensitivity was found for point-of-care tests of whole blood and serum or plasma (98.9% [95% CI, 94.5% to 99.8%] and 98.9% [95% CI, 96.8% to 99.6%], respectively), while sensitivity for rapid diagnostic tests of serum or plasma and point-of-care tests for oral fluid were 98.4% (95% CI, 88.9% to 99.8%) and 97.1% (95% CI, 94.7% to 98.4%), respectively. Point-of-care tests of whole blood and serum or plasma had the highest specificity (99.5% [95% CI, 97.5% to 99.9%] and 99.7% [95% CI, 99.3% to 99.9%], respectively), while rapid diagnostic tests of serum or plasma and point-of-care tests of oral fluid had specificities of 98.6% (95% CI, 94.9% to 99.6%) and 98.2% (95% CI, 92.2% to 99.6%), respectively.

The authors noted that test accuracy could have been affected by the use of different reference standards and by co-infection and that the tests used could not differentiate between chronic and acute infection, among other limitations. However, they concluded that point-of-care tests of blood and oral fluid and rapid diagnostic tests of serum or plasma are accurate enough to be used for HCV screening.

“In light of their accuracy and the urgent need to increase hepatitis C screening in marginalized and at-risk populations and in endemic HCV settings, these tests could play a substantial role in expanded global screening initiatives, which would eventually impact the control of HCV infection at the population level,” they wrote.

Report highlights care delivery differences at top hospitals

A recent Dartmouth Atlas Project report aims to capture differences in the way top-rated academic hospitals deliver care, so that residents can make informed choices about where they train.

The report was drawn from 2010 Medicare data about medical and surgical care at 23 notable academic medical centers, including those rated by U.S. News and World Report as the best hospitals for clinical excellence. It examines variations in the use of medical resources to treat patients at the end of life, as well as quality-of-care metrics about processes of care, patient experience and patient safety. The hospitals in the report held about 17% of all primary residency slots in 2012.

In the area of end-of-life care, for example, 49% of chronically ill patients treated at Johns Hopkins Hospital in Baltimore were enrolled in hospice in the last six months of their lives, compared to 23% of patients at New York's Mt. Sinai Medical Center. “A Mount Sinai resident may therefore learn a higher threshold for referral to hospice care, or decide to explore more aggressive treatment first,” a press release said. “A Johns Hopkins resident may be better trained in having discussions with patients about their preferences.”

Other differences between hospitals include the following:

  • Sixty-seven percent of chronically ill patients at NYU Langone Medical Center in New York City saw 10 or more physicians in their last six months of life, compared to 43% of patients at Scott & White Memorial Hospital in Temple, Texas.
  • Patients at NYU Langone Medical Center were 47 times less likely to get a urinary catheter infection than patients at the University of Michigan Health System in Ann Arbor, Mich.
  • Patients were twice as likely to get knee replacement surgery in Salt Lake City (11.9 per 1,000 discharges) as in Manhattan (4.5 per 1,000 discharges).

“During their residency training, young physicians learn by observing faculty, making decisions on how aggressively to treat chronically ill patients at the end of life, and whether to recommend surgery when other treatment options exist,” said John R. Lumpkin, MD, MPH, director of the Health Care Group at the Robert Wood Johnson Foundation, a longtime funder of the Dartmouth Atlas Project, in the release. “These variations in the way care is delivered are not trivial, as they may very well affect the future practice of medicine.”

QI efforts to enhance communication with very ill can affect outcomes, analysis finds

Quality improvement (QI) interventions to improve communication with very sick patients are effective at reducing length of stay in the ICU and the use of nonbeneficial aggressive treatment, a review found.

Researchers used several data sources, such as MEDLINE and Cochrane, to systematically review 20 studies on QI interventions aimed at improving communication with patients with life-limiting or severe life-threatening illness and/or their families. They also sought to determine whether a consultative or integrative intervention approach worked best. The former involves consulting with communication specialists such as palliative medicine teams, while the latter involves embedding palliative care principles and intervention into the practice of the usual care team. Researchers included only prospective, controlled studies published between January 2000 and December 2011 in their analysis.

Thirteen (65%) of the studies took place in the intensive care unit (ICU). The four main intervention types, and their associated rates of improvement in health care utilization (like lower ICU stay length), were:

  • family meetings with the usual team (11 studies, 77% of which saw improvement in utilization),
  • palliative care teams (5 studies, 50% saw improvement),
  • ethics consultation (2 studies, 100% saw improvement) and
  • physician-patient communication (2 studies, no improvement).

Among those studies that addressed health care utilization, such as ICU length of stay and potentially nonbeneficial life-sustaining aggressive treatment, 73% found statistically significant improvement. In studies that evaluated intervention impact on quality of life, symptom control and satisfaction, most findings weren't significant. A higher proportion of consultative studies found statistically significant improvement related to the intervention than did integrative studies. Results were published in October 2012 by the Journal of General Internal Medicine.

“Our study supports that consultative interventions, as opposed to integrative ones, are more likely to potentially improve healthcare utilization, although there are successful examples of both interventions types,” the authors wrote. This may be, in part, because consultative teams can focus more on communication without the need to balance clinical duties, they noted.

DNR orders not associated with care quality in stroke survivors

Do-not-resuscitate (DNR) orders were not associated with quality of care after acute ischemic stroke, a study found.

Researchers abstracted chart data from patients admitted to 131 Veterans Health Administration facilities in fiscal year 2007 to determine whether there was a link between DNR orders and quality of care after stroke. A patient was considered to have a DNR order if an electronic order was present or if “no code,” “no cardiopulmonary resuscitation,” or “no resuscitation” was documented in the chart. Quality of care was determined by 14 indicators of quality of ischemic stroke care for inpatients. Multivariable logistic regression was used to examine any relationship between a DNR order and care quality. The study was published in the Nov. 6, 2012, Neurology.

The 14 quality indicators were divided into early hospital measures, in-hospital measures, and discharge-related measures. Among the 3,695 patients included in the study, 535 (13.5%) had DNR orders, and of these, 71% were initially documented within a day of hospital admission. Overall, 4.9% of patients died in the hospital or were discharged to hospice. These outcomes were more common among patients with DNR orders (29.7%), particularly those with documentation two or more days after admission (47.1%). DNR orders were more common in older patients, patients with more comorbid conditions, and patients with more severe strokes. Early ambulation was less likely in patients with DNR orders (odds ratio, 0.58; 95% CI, 0.41 to 0.81), but no significant association was seen for other quality indicators.

The authors noted that the clinical context in which the DNR orders were placed was not known and that the prevalence of DNR orders varied widely among facilities, among other limitations. However, they concluded that while patients in this population who had DNR orders were more likely to have poor outcomes, no association was seen between DNR orders and quality of care. They called for further studies to examine the quality of DNR discussions in veterans' facilities, the policies that cause interfacility variation in DNR documentation, clinical and other factors associated with DNR orders, and “the impact of DNR orders on the full range of potentially life-sustaining treatments and interventions.” Whether the relationship between DNR orders and quality of care differs among hospitals should also be studied, the authors wrote.