In the News

Mortality disparities after in-hospital cardiac arrest, and more.


Blacks more likely than whites to die after in-hospital cardiac arrest

Survival rates are worse for black patients admitted to the hospital after cardiac arrest than for whites, largely because blacks are more likely to be treated at low-performing hospitals, a recent study found.

Researchers studied a cohort of 10,011 patients (18.8% black; 81.2% white) with cardiac arrests due to ventricular fibrillation or pulseless ventricular tachycardia at 274 hospitals between 2000 and 2008. Black patients had a 27% lower relative risk and a 12% lower absolute rate of survival to hospital discharge than whites. The disparities were largely due to blacks being more likely than whites to be admitted to hospitals with worse outcomes, the authors said, as well as to differences in quality of care during the acute and post-resuscitation phases. The study appeared in the Sept. 16 Journal of the American Medical Association.

The results suggest that black patients are more likely to have cardiac arrests in hospitals with higher rates of delays in defibrillation time, the authors said. They noted physician bias does not explain racial differences during the acute resuscitation phase because they found no evidence of differences in aggressiveness of resuscitation, such as the number of attempted defibrillations and total treatment time.

In the post-resuscitation phase, the hospital itself accounted for most of the racial differences, they said. Further study is needed to determine whether the lower rates of post-resuscitation survival in blacks are due to specific hospital characteristics, such as lower-quality ICU expertise or less aggressive use of therapies such as hypothermia and cardiac catheterization, the authors said.

The authors concluded that strategies to eliminate racial disparities must involve improving resuscitation survival and post-resuscitation care in poor-performing hospitals that serve large populations of black patients.

More patients leaving hospital against medical advice

The percentage of patients who left the hospital against medical advice increased significantly between 1997 and 2007, according to a recent brief from the Agency for Healthcare Research and Quality.

In 2007, hospitalizations that ended in patients leaving against medical advice (AMA) accounted for 368,000 hospital stays (1.2% of the total) compared to only 264,000 discharges in 1997, a 39% increase. Certain characteristics also distinguished the AMA stays. Patients were more likely to be young (average age 46) and male. Overall women are slightly more likely to be hospitalized, but men left AMA 60% more often than women. Three of the five most common diagnoses for patients who left AMA were related to mental health and substance abuse. Nonspecific chest pain and diabetes with complications were the other two top diagnoses.

Patients were also more likely to leave if they had either Medicaid coverage or no insurance at all. On average, AMA stays were about half as long and half as expensive as other hospitalizations. There were also differences by geographic area. The Northeast had double the rate of AMA departures, while the West had the lowest rate. Patients living in large urban areas were nearly twice as likely to leave AMA compared with patients living in all other areas.

The statistical brief was based on data from the Healthcare Cost and Utilization Project 2007 Nationwide Inpatient Sample. Patients who leave the hospital AMA have higher readmission rates and may be at increased risk for adverse health outcomes, noted the AHRQ report. Greater understanding of the characteristics of these AMA stays may therefore assist in the design of strategies to prevent patients from leaving the hospital prematurely.

Guideline-compliant therapy produces better outcomes for hospitalized CAP patients

Two recent studies in Archives of Internal Medicine support treating hospitalized community-acquired pneumonia (CAP) patients according to recommendations in current clinical practice guidelines.

In one study, researchers evaluated the association between in-hospital survival and guideline-concordant therapy in more than 54,000 non-intensive care unit inpatients with CAP at 113 community hospitals and tertiary care centers. After controlling for severity of illness and other patient characteristics, they found a significant decrease in in-hospital mortality (odds ratio, 0.70; 95% CI, 0.63 to 0.77) and a 0.6-day decrease in length of stay (LOS) in patients treated according to guidelines. Guideline-compliant treatment also was associated with fewer complications, such as sepsis and renal failure, and earlier switch to oral therapy.

Improved outcomes were linked to the use of fluoroquinolones or macrolides, which provide coverage for atypical organisms, the authors noted, and were associated with a Pneumonia Severity Index-adjusted reduction in mortality of 20% to 40% compared with treatment regimens that excluded these antibiotic classes. Until further data are available, the study supports the use of the Infectious Diseases Society of America (IDSA)/American Thoracic Society (ATS) guidelines as the default treatment plan for non-ICU hospitalized adults with CAP, the authors concluded.

In a second study, researchers analyzed a database of 1,649 patients age 65 or older who were hospitalized with CAP at 43 centers in 12 countries. Of the patients treated according to guidelines, 71% reached clinical stability within seven days, compared with 57% in the nonadherent group. Adherence to guidelines also was associated with shorter LOS and decreased in-hospital mortality. The absolute risk reduction in mortality was 9.9% with a number needed to treat of 10. The authors theorized that the benefit of using fluoroquinolones or adding macrolides to beta-lactum CAP regimens may be related to targeted therapy for atypical pathogens, especially Legionella species.

An accompanying editorial noted that future studies of CAP should focus on rapid bedside testing to identify patients most likely to benefit from guideline-concordant therapy.

These two articles add to a growing body of evidence supporting guideline-compliant therapy for hospitalized patients with CAP, the editorial said. While further research is needed to confirm the results, the editorialist added, the evidence is compelling enough to support guideline-concordant antibiotic regimens as the default treatment for typical and atypical organisms. The editorialist added that hospitals should standardize treatment for all CAP patients based on current guidelines.

8.3 million adults seriously considered suicide last year

Nearly 8.3 million U.S. adults (3.7%) had serious thoughts of committing suicide in the past year, reported the Substance Abuse and Mental Health Services Administration, a division of the Department of Health and Human Services.

The report is based on 2008 data drawn from the National Survey on Drug Use and Health, which obtained responses from 46,190 persons aged 18 years or older. The study also showed that 2.3 million adults made a suicide plan, 1.1 million adults attempted suicide, and adult women had marginally higher levels of suicidal thoughts and behaviors than males.

People experiencing substance abuse disorders within the past year were more than three times as likely to have seriously considered suicide (11% vs. 3% of non-substance abusers). Substance abusers were also four times more likely to have planned a suicide than those without substance abuse disorders (3.4% vs. 0.8%), and nearly seven times more likely to have attempted suicide (2% vs. 0.3%).

Of those adults who attempted suicide in the past year, 62.3% received medical attention for their attempts, and 46% stayed in a hospital overnight or longer for treatment.

Localization of physicians improves communication with nurses

Assigning physicians to a specific unit within the hospital improved communication and understanding with nurses, but did not improve agreement on all aspects of the plan of care, according to a new study.

The study was conducted in an academic medical center that switched from having patients assigned to a physician service and then randomly admitted to a unit to a system in which physician teams were geographically localized to specific units in the hospital. Three units were designated for the teaching service, and four units were covered by hospitalists. Researchers interviewed about 300 nurses and 300 physicians before and after the change. The results were published by the Journal of General Internal Medicine on Sept. 19.

After the localization was put in place, physicians were significantly more likely to be able to identify the nurse who was treating a particular patient (58% vs. 36%, P<0.001). The same was true for nurses' ability to correctly identify physicians (93% vs. 71%, P<0.001). Higher percentages of both groups also reported having communicated with each other, and more communication occurred face to face instead of by phone.

The study also looked at agreement between nurses and physicians on patients' plans of care, and found improved cohesion on anticipated length of stay (48% in complete agreement after vs. 33% before) and planned tests (69% vs. 59%) but no significant improvement in other areas.

Researchers concluded that localization can improve several aspects of nurse-physician communication. However, they noted, proximity is not sufficient to create a shared understanding among team members on the plan of care, although proximity did allow the researchers to implement interventions such as interdisciplinary rounds. The researchers suggested that additional techniques, such as team training, be investigated as means to further improve communication.

Hospital's census reduction strategy offers disaster planning insight

Cancelling elective surgeries and admissions and expediting discharges before a planned relocation helped one hospital decrease its census by 36% in one week, and its experience may be applicable to disaster planning, according to a new study.

During a planned move of operations at UCLA Medical Center to a new facility, researchers performed a prospective analysis of the hospital's activities to help identify ways to improve surge capacity in the event of a disaster. The researchers analyzed census data on the hospital's operations two weeks before the planned move, or baseline, and one week before the planned move, or the transition period. The study results appear in the September Archives of Surgery.

The medical center's census management strategy used the following three components

  • restricting the elective surgery schedule starting one week before the planned move,
  • limiting incoming transfers, and
  • delegating discharges to a multidisciplinary team to increase efficiency.

At baseline, the average daily census was 537 patients. The census was reduced by a rate of 18 patients per day during the transition period to 345 patients on the day of the move, a reduction of 36%. All services saw decreases, but reductions were larger for surgical than for nonsurgical services (46% vs. 30%; P=0.02). The ICU and medical/surgical wards both saw significant reductions in their censuses, although the latter reduction was greater (17% vs. 40%; P<0.001). Surgery volume decreased by 45% between the baseline and transition periods, entirely due to reductions in elective surgeries; rates of emergency surgeries did not change. Admissions decreased by 42%, while an 8% increase was seen in adjusted discharges per occupied bed. No effect on inpatient mortality was noted.

The authors concluded that the strategy used to decrease UCLA Medical Center's census before its planned move could also be used in planning for disasters with longer lead times, such as hurricanes or pandemics. They noted that it took three or four days to achieve a significant decrease in the census and a resulting availability of acute care beds. “This strategy might serve as a model both for large-scale disaster inpatient surge capacity planning as a component of hospital disaster preparedness and for day-to-day census management in individual hospitals,” they wrote.

AHRQ offers free tools for disaster planning

The Agency for Healthcare Research and Quality has released two free tools to assist hospitals in disaster planning.

The interactive computer-based tools will help emergency responders and planners select and equip alternate care sites and determine which patients can be moved to these sites. The Disaster Alternate Care Facilities Selection Tool allows users to evaluate potential sites, such as college campuses, schools or convention centers, and determine what would need to be done to prepare them for use in a disaster. The Disaster Alternate Care Facility Patient Selection Tool helps clinicians match hospitalized patients with a potential alternate care facility, freeing up space in the hospital for incoming patients.

The tools are available free of charge online.< /p>

SHM begins new glucose control project

The launch of a new glycemic control program was announced recently by the Society of Hospital Medicine (SHM).

The Glycemic Control Mentored Implementation (GCMI) is intended to improve early detection and treatment of hyper- or hypoglycemia in hospitalized patients. By addressing pertinent clinical and systems issues, GCMI aims to improve care of patients with diabetes and to reduce its associated complications. The effort, which is supported by funding from sanofi-aventis, uses mentors who are glycemic control and quality improvement experts to provide direct support to hospitalist-led teams.

Each team enrolled in the two-year project receives a dedicated mentor, as well as Web-based and print resources on glycemic control. The project will address the use of both subcutaneous and intravenous insulin, the two most common methods for managing glycemic levels in the hospital.

Specific topics include

  • access to subcutaneous insulin on the general medical/ surgical floor,
  • intravenous insulin infusion in the intensive care unit,
  • transitioning patients from intravenous to subcutaneous insulin, and
  • transitioning patients home on insulin therapy.

The project will include 30 hospitals in 22 states.

Three-point prediction model may help diagnose AHF in the ED

A prediction model that combines biomarkers and clinical judgment may improve diagnosis of acute heart failure in the emergency department, a study concluded.

In the study, physician estimates of probability of acute heart failure (AHF) in 500 patients treated in the emergency department were taken from the Improved Management of Patients with Congestive Heart Failure trial. A model using the variables of age, pre-test probability and log N-terminal pro-B-type natriuretic peptide (NT-proBNP) correctly reclassified 44% of patients initially estimated at intermediate probability of AHF to either low or high probability. The study was published in the Oct. 13 Journal of the American College of Cardiology.

Findings suggest that interpreting biomarkers such as NT-proBNP as continuous variables, as opposed to categorical tests, is a superior approach, the authors said. The strategy may be particularly useful in patients for whom the diagnosis is unclear and who often need ancillary tests. In the study, use of the model led almost half of patients classified as intermediate risk to be redirected, with 99% accuracy, the authors said.

The study confirms many physicians' belief that clinical judgment must be incorporated into prediction rules, the authors continued. This model is easy to use because it doesn't require elaborate clinical information and can quickly provide guidance when a physician is undecided about an AHF diagnosis, they said.

Mortality rates improve in U.S. hospitals, but quality gaps persist

Mortality rates in U.S. hospitals improved from 2006 through 2008, but care quality continues to vary, according to a new study.

In its 12th annual Hospital Quality in America Study, health care ratings company HealthGrades analyzed data on approximately 40 million Medicare discharges to measure variations in quality of care. Researchers found that although U.S. in-hospital risk-adjusted mortality rates improved an average of 10.99% from 2006 to 2008, quality of care improved more quickly at highly rated (five-star) hospitals than at lower rated (one- or three-star) hospitals (11.89% vs. 10.14% or 10.72%, respectively).

Large gaps in quality between hospitals were also observed, the researchers noted. For all procedures and diagnoses studied, patients cared for in a five-star hospital had a 71.64% lower risk for death than those cared for in a one-star hospital. Of 224,537 potentially preventable Medicare deaths, approximately 56% (127,488) were associated with four diagnoses

  • sepsis (44,622 deaths),
  • pneumonia (29,251 deaths),
  • heart failure (26,374 deaths) and
  • respiratory failure (27,241 deaths)

The full HealthGrades report is available online.

In the News is a product of ACP HospitalistWeekly , an e-newsletter provided every Wednesday by ACP Hospitalist . If you're not already receiving ACP HospitalistWeekly , contact Customer Service at 800-523-1546, ext. 2600, or directly at 215-351-2600 (M-F, 9 a.m. to 5 p.m. EST) or send an email.