Mary Jo M. Dales

Editor Robin Turner contributed to this story.

Medicare costs for chronic disease care at the end of life differed nearly twofold among the nation's “top five” academic medical centers as rated by U.S. News & World Report.

The finding, derived from data in the Dartmouth Atlas of Health Care 2008 report “Tracking the Care of Patients with Severe Chronic Illness,” convincingly argues that providing more services isn't necessary for providing good care to these patients, according to the report's authors.

Their broader analysis examined 2001–2005 data from Medicare's 306 hospital referral regions, and was limited to services provided in the last 2 years of life for all enrollees who had any of nine specific severe and chronic diseases.

The leading chronic diseases included: chronic heart failure, cancer, and chronic obstructive pulmonary disease. Costs and services were evaluated for inpatient hospital care, outpatient services, skilled nursing and long-term hospital care, home health care, and hospice.

A high supply of available health care services encouraged more use of services, Dr. John E. Wennberg and the coauthors of the report found.

Paradoxically, the highest volume of services was perceived to be poorer care by both patients and physicians. The researchers attributed this inverse relationship to increased chances for errors, complications, miscommunications, and confusion about physician responsibilities for care as the volume of services increases.

Overall, 55% of total costs were for acute hospital care. The next largest fraction of spending (nearly 16%) was for outpatient care.

The availability of alternative care—skilled nursing, rehab facilities, home health care, and hospices—did not necessarily lead to declines in hospitalizations or inpatient spending. Instead, admission to hospitals increased demand for discharge to other care sectors.

In regions with the most supply-sensitive care, patients spent nearly 22 days as inpatients and averaged almost 60 physician visits during their last 6 months of life. In areas with the least supply-sensitive care, patients spent about 6 days in hospitals and averaged 15 physician visits in their last 6 months.

To address the relationship between quality of care and cost of care, the authors took the novel approach of comparing Medicare spending and the availability and utilization of resources at the nation's top five academic medical centers (see box, next page). At the University of California, Los Angeles, Medical Center, Medicare spent more than $93,000 per chronically ill patient in the last 2 years of life.

At the Mayo Clinic, Rochester, Minn., and the Cleveland Clinic, however, the costs were about $53,000 and $55,000 respectively. Johns Hopkins Hospital, Baltimore, and Massachusetts General Hospital, Boston, ranked second and third in total Medicare spending.

As in the larger analysis, the differences in spending at the top five centers were mainly driven by the supply of services. Compared with the other four centers, UCLA Medical Center has proportionately the highest numbers of physicians, hospital beds, ICU beds, and imaging and other services. Again, most of the differences in cost were in acute care; for example, 50% more days spent in the hospital in the last 6 months of life by patients at UCLA (18.5 days), compared with those at the Mayo Clinic (12 days).

The authors credited the group practice model and coordination of care at the Cleveland Clinic and Mayo Clinic with the cost savings, and proposed that other systems emulate the spending, resource input, and utilization profiles of the relatively efficient clinics.

Benchmarking based on practices at efficient centers can then be used to calculate potential savings by reducing overuse of supply-sensitive services at high-cost centers.

The Dartmouth findings come as no surprise to emergency physicians, said Dr. Brent Asplin, head of emergency medicine at Regions Hospital and HealthPartners Medical Group in St. Paul, Minn.

“The emergency department has a view of what's going on in the broader health care system. We see the revolving door of repeat hospitalizations and emergency department visits for patients with poorly managed chronic disease.

“When you reward providers based on the volume of services, you will see higher utilization in areas with more providers. We need to work toward a system that rewards value rather than volume,” he said.

“One of the keys for addressing the wide variation in services is to restructure the way primary care is organized so there is a coordinated team to manage patients with chronic disease,” Dr. Asplin said.

“Ultimately, we cannot hospitalize our way to better health,” Dr. Asplin said, adding “It is much easier to send a patient with chronic disease home from the ED when you know there is a primary care team that will pick up that patient the next day.”

Dr. Frank Michota, founder of the Cleveland Clinic's hospital medicine program, said the hospital strives for appropriate utilization by identifying “the hospitalization goal for each chronically ill patient who is admitted and driving the care plan to meet specific objectives.”

“We have no illusions that we will make a chronically ill patient normal again, but our default position is to treat aggressively until it is clear that no reversible pathology exists,” Dr. Michota commented.

“Full discussion with the patient or family on the feasibility or likelihood of achieving the goal is also important,” he said. Medically futile care plans are not undertaken just because that is what the patient or family wants.

The report authors recommended research on how treatments affect outcomes, patients' lives, and the efficiency of clinical practice.

Evidence is lacking for how often to see patients, when to refer to specialists, and when to admit.

As a result, primary care physicians will refer to a specialist or admit to a hospital if those resources are available and payments for office-based care are constrained, they said.

Patients need to be followed over time and across settings by established group practices and integrated provider systems that are capable of organizing care over the span of an individual's chronic illness. Organizations that participate in this research should be rewarded through a proposed shared-savings program with the Centers for Medicare and Medicaid Services that is designed to encourage coordination and to reduce overuse of care, they proposed.

Physician groups and hospitals should be encouraged to become real or virtual integrated systems that are willing to be accountable for the coordination, overall costs, and quality of care provided to chronic disease patients.

The authors proposed a shared-savings approach in which payments are based on per-beneficiary costs relative to appropriate spending targets.

Shared savings would allow physicians and hospitals to preserve their net incomes while reducing total revenues resulting from unnecessary care and overuse of acute care hospitals.

In addition to the Robert Wood Johnson Foundation, other supporters of the Dartmouth Atlas project include the WellPoint Foundation, Aetna Foundation, United Health Foundation, and California Healthcare Foundation. The full report is available at www.dartmouthatlas.org

ELSEVIER GLOBAL MEDICAL NEWS

Editor Robin Turner contributed to this story.

Medicare costs for chronic disease care at the end of life differed nearly twofold among the nation's “top five” academic medical centers as rated by U.S. News & World Report.

The finding, derived from data in the Dartmouth Atlas of Health Care 2008 report “Tracking the Care of Patients with Severe Chronic Illness,” convincingly argues that providing more services isn't necessary for providing good care to these patients, according to the report's authors.

Their broader analysis examined 2001–2005 data from Medicare's 306 hospital referral regions, and was limited to services provided in the last 2 years of life for all enrollees who had any of nine specific severe and chronic diseases.

The leading chronic diseases included: chronic heart failure, cancer, and chronic obstructive pulmonary disease. Costs and services were evaluated for inpatient hospital care, outpatient services, skilled nursing and long-term hospital care, home health care, and hospice.

A high supply of available health care services encouraged more use of services, Dr. John E. Wennberg and the coauthors of the report found.

Paradoxically, the highest volume of services was perceived to be poorer care by both patients and physicians. The researchers attributed this inverse relationship to increased chances for errors, complications, miscommunications, and confusion about physician responsibilities for care as the volume of services increases.

Overall, 55% of total costs were for acute hospital care. The next largest fraction of spending (nearly 16%) was for outpatient care.

The availability of alternative care—skilled nursing, rehab facilities, home health care, and hospices—did not necessarily lead to declines in hospitalizations or inpatient spending. Instead, admission to hospitals increased demand for discharge to other care sectors.

In regions with the most supply-sensitive care, patients spent nearly 22 days as inpatients and averaged almost 60 physician visits during their last 6 months of life. In areas with the least supply-sensitive care, patients spent about 6 days in hospitals and averaged 15 physician visits in their last 6 months.

To address the relationship between quality of care and cost of care, the authors took the novel approach of comparing Medicare spending and the availability and utilization of resources at the nation's top five academic medical centers (see box, next page). At the University of California, Los Angeles, Medical Center, Medicare spent more than $93,000 per chronically ill patient in the last 2 years of life.

At the Mayo Clinic, Rochester, Minn., and the Cleveland Clinic, however, the costs were about $53,000 and $55,000 respectively. Johns Hopkins Hospital, Baltimore, and Massachusetts General Hospital, Boston, ranked second and third in total Medicare spending.

As in the larger analysis, the differences in spending at the top five centers were mainly driven by the supply of services. Compared with the other four centers, UCLA Medical Center has proportionately the highest numbers of physicians, hospital beds, ICU beds, and imaging and other services. Again, most of the differences in cost were in acute care; for example, 50% more days spent in the hospital in the last 6 months of life by patients at UCLA (18.5 days), compared with those at the Mayo Clinic (12 days).

The authors credited the group practice model and coordination of care at the Cleveland Clinic and Mayo Clinic with the cost savings, and proposed that other systems emulate the spending, resource input, and utilization profiles of the relatively efficient clinics.

Benchmarking based on practices at efficient centers can then be used to calculate potential savings by reducing overuse of supply-sensitive services at high-cost centers.

The Dartmouth findings come as no surprise to emergency physicians, said Dr. Brent Asplin, head of emergency medicine at Regions Hospital and HealthPartners Medical Group in St. Paul, Minn.

“The emergency department has a view of what's going on in the broader health care system. We see the revolving door of repeat hospitalizations and emergency department visits for patients with poorly managed chronic disease.

“When you reward providers based on the volume of services, you will see higher utilization in areas with more providers. We need to work toward a system that rewards value rather than volume,” he said.

“One of the keys for addressing the wide variation in services is to restructure the way primary care is organized so there is a coordinated team to manage patients with chronic disease,” Dr. Asplin said.

“Ultimately, we cannot hospitalize our way to better health,” Dr. Asplin said, adding “It is much easier to send a patient with chronic disease home from the ED when you know there is a primary care team that will pick up that patient the next day.”

Dr. Frank Michota, founder of the Cleveland Clinic's hospital medicine program, said the hospital strives for appropriate utilization by identifying “the hospitalization goal for each chronically ill patient who is admitted and driving the care plan to meet specific objectives.”

“We have no illusions that we will make a chronically ill patient normal again, but our default position is to treat aggressively until it is clear that no reversible pathology exists,” Dr. Michota commented.

“Full discussion with the patient or family on the feasibility or likelihood of achieving the goal is also important,” he said. Medically futile care plans are not undertaken just because that is what the patient or family wants.

The report authors recommended research on how treatments affect outcomes, patients' lives, and the efficiency of clinical practice.

Evidence is lacking for how often to see patients, when to refer to specialists, and when to admit.

As a result, primary care physicians will refer to a specialist or admit to a hospital if those resources are available and payments for office-based care are constrained, they said.

Patients need to be followed over time and across settings by established group practices and integrated provider systems that are capable of organizing care over the span of an individual's chronic illness. Organizations that participate in this research should be rewarded through a proposed shared-savings program with the Centers for Medicare and Medicaid Services that is designed to encourage coordination and to reduce overuse of care, they proposed.

Physician groups and hospitals should be encouraged to become real or virtual integrated systems that are willing to be accountable for the coordination, overall costs, and quality of care provided to chronic disease patients.

The authors proposed a shared-savings approach in which payments are based on per-beneficiary costs relative to appropriate spending targets.

Shared savings would allow physicians and hospitals to preserve their net incomes while reducing total revenues resulting from unnecessary care and overuse of acute care hospitals.

In addition to the Robert Wood Johnson Foundation, other supporters of the Dartmouth Atlas project include the WellPoint Foundation, Aetna Foundation, United Health Foundation, and California Healthcare Foundation. The full report is available at www.dartmouthatlas.org

ELSEVIER GLOBAL MEDICAL NEWS

Editor Robin Turner contributed to this story.

Medicare costs for chronic disease care at the end of life differed nearly twofold among the nation's “top five” academic medical centers as rated by U.S. News & World Report.

The finding, derived from data in the Dartmouth Atlas of Health Care 2008 report “Tracking the Care of Patients with Severe Chronic Illness,” convincingly argues that providing more services isn't necessary for providing good care to these patients, according to the report's authors.

Their broader analysis examined 2001–2005 data from Medicare's 306 hospital referral regions, and was limited to services provided in the last 2 years of life for all enrollees who had any of nine specific severe and chronic diseases.

The leading chronic diseases included: chronic heart failure, cancer, and chronic obstructive pulmonary disease. Costs and services were evaluated for inpatient hospital care, outpatient services, skilled nursing and long-term hospital care, home health care, and hospice.

A high supply of available health care services encouraged more use of services, Dr. John E. Wennberg and the coauthors of the report found.

Paradoxically, the highest volume of services was perceived to be poorer care by both patients and physicians. The researchers attributed this inverse relationship to increased chances for errors, complications, miscommunications, and confusion about physician responsibilities for care as the volume of services increases.

Overall, 55% of total costs were for acute hospital care. The next largest fraction of spending (nearly 16%) was for outpatient care.

The availability of alternative care—skilled nursing, rehab facilities, home health care, and hospices—did not necessarily lead to declines in hospitalizations or inpatient spending. Instead, admission to hospitals increased demand for discharge to other care sectors.

In regions with the most supply-sensitive care, patients spent nearly 22 days as inpatients and averaged almost 60 physician visits during their last 6 months of life. In areas with the least supply-sensitive care, patients spent about 6 days in hospitals and averaged 15 physician visits in their last 6 months.

To address the relationship between quality of care and cost of care, the authors took the novel approach of comparing Medicare spending and the availability and utilization of resources at the nation's top five academic medical centers (see box, next page). At the University of California, Los Angeles, Medical Center, Medicare spent more than $93,000 per chronically ill patient in the last 2 years of life.

At the Mayo Clinic, Rochester, Minn., and the Cleveland Clinic, however, the costs were about $53,000 and $55,000 respectively. Johns Hopkins Hospital, Baltimore, and Massachusetts General Hospital, Boston, ranked second and third in total Medicare spending.

As in the larger analysis, the differences in spending at the top five centers were mainly driven by the supply of services. Compared with the other four centers, UCLA Medical Center has proportionately the highest numbers of physicians, hospital beds, ICU beds, and imaging and other services. Again, most of the differences in cost were in acute care; for example, 50% more days spent in the hospital in the last 6 months of life by patients at UCLA (18.5 days), compared with those at the Mayo Clinic (12 days).

The authors credited the group practice model and coordination of care at the Cleveland Clinic and Mayo Clinic with the cost savings, and proposed that other systems emulate the spending, resource input, and utilization profiles of the relatively efficient clinics.

Benchmarking based on practices at efficient centers can then be used to calculate potential savings by reducing overuse of supply-sensitive services at high-cost centers.

The Dartmouth findings come as no surprise to emergency physicians, said Dr. Brent Asplin, head of emergency medicine at Regions Hospital and HealthPartners Medical Group in St. Paul, Minn.

“The emergency department has a view of what's going on in the broader health care system. We see the revolving door of repeat hospitalizations and emergency department visits for patients with poorly managed chronic disease.

“When you reward providers based on the volume of services, you will see higher utilization in areas with more providers. We need to work toward a system that rewards value rather than volume,” he said.

“One of the keys for addressing the wide variation in services is to restructure the way primary care is organized so there is a coordinated team to manage patients with chronic disease,” Dr. Asplin said.

“Ultimately, we cannot hospitalize our way to better health,” Dr. Asplin said, adding “It is much easier to send a patient with chronic disease home from the ED when you know there is a primary care team that will pick up that patient the next day.”

Dr. Frank Michota, founder of the Cleveland Clinic's hospital medicine program, said the hospital strives for appropriate utilization by identifying “the hospitalization goal for each chronically ill patient who is admitted and driving the care plan to meet specific objectives.”

“We have no illusions that we will make a chronically ill patient normal again, but our default position is to treat aggressively until it is clear that no reversible pathology exists,” Dr. Michota commented.

“Full discussion with the patient or family on the feasibility or likelihood of achieving the goal is also important,” he said. Medically futile care plans are not undertaken just because that is what the patient or family wants.

The report authors recommended research on how treatments affect outcomes, patients' lives, and the efficiency of clinical practice.

Evidence is lacking for how often to see patients, when to refer to specialists, and when to admit.

As a result, primary care physicians will refer to a specialist or admit to a hospital if those resources are available and payments for office-based care are constrained, they said.

Patients need to be followed over time and across settings by established group practices and integrated provider systems that are capable of organizing care over the span of an individual's chronic illness. Organizations that participate in this research should be rewarded through a proposed shared-savings program with the Centers for Medicare and Medicaid Services that is designed to encourage coordination and to reduce overuse of care, they proposed.

Physician groups and hospitals should be encouraged to become real or virtual integrated systems that are willing to be accountable for the coordination, overall costs, and quality of care provided to chronic disease patients.

The authors proposed a shared-savings approach in which payments are based on per-beneficiary costs relative to appropriate spending targets.

Shared savings would allow physicians and hospitals to preserve their net incomes while reducing total revenues resulting from unnecessary care and overuse of acute care hospitals.

In addition to the Robert Wood Johnson Foundation, other supporters of the Dartmouth Atlas project include the WellPoint Foundation, Aetna Foundation, United Health Foundation, and California Healthcare Foundation. The full report is available at www.dartmouthatlas.org

ELSEVIER GLOBAL MEDICAL NEWS

SAN DIEGO — Recent trial results have quelled much of the excitement about the promise of interventions designed to reduce the risk of death resulting from sepsis. Yet within the mostly negative results of those studies, there are some findings that help to define appropriate treatment approaches, according to Dr. David A. Schulman, chief of pulmonary and critical care medicine at Emory University Hospital, Atlanta.

The findings of VASST (Vasopressin and Septic Shock Trial) found no difference in 28-day mortality among 778 patients randomized to norepinephrine or to norepinephrine plus vasopressin. Although the difference was not statistically significant, those who had less severe shock and were started on less than 15 mcg/min of norepinephrine had a lower mortality rate with the addition of vasopressin (26.5%), compared with the rate seen in those given norepinephrine alone (35.7%). The result persisted at 90 days (N. Engl. J. Med. 2008;358:877–87).

Thus, vasopressin can be used as a second agent in the hypotensive, septic patient who is on a moderate dose of another pressor, Dr. Schulman said in a presentation at the annual meeting of the Society of Hospital Medicine. Vasopressin, dosed at 0.03 U/min, may be added to norepinephrine with an anticipated effect equivalent to that of norepinephrine alone. Vasopressin should not be titrated, and it should be used only as a second agent. Vasopressin should be used with caution in patients with significant myocardial dysfunction (cardiac index less than 2.1 L/min per m

The recent CORTICUS (Corticosteroid Therapy of Septic Shock) trial concluded that steroids are no better than placebo for reducing mortality, yet the role of steroid therapy in the treatment of septic shock is “more unclear than ever,” Dr. Schulman said. Current data do not provide sufficient evidence that low doses of corticosteroids are harmful in septic patients without relative adrenal insufficiency. The current advisory is to “implement a combination of hydrocortisone and fludrocortisone if a patient remains hypotensive with presumed septic shock for [more than 1 hour] after administration of appropriate fluids and pressors.”

CORTICUS, a randomized controlled trial of 500 septic shock patients, found that 28-day mortality was comparable whether subjects were given placebo or hydrocortisone at 50 mg every 6 hours for 5 days followed by a 6-day taper. Furthermore, the trial discerned no difference in response to steroids among patients with relative adrenal insufficiency. Shock resolved faster in the steroid-treated subjects, but they didn't live any longer than patients who were not given steroids. Also, steroids were associated with an increased risk of hyperglycemia, recurrent sepsis, and recurrent shock (N. Engl. J. Med. 2008;358:111–24).

The findings countered those of the practice-changing study that held that adrenal insufficiency predicted which vasopressor-refractory sepsis patients would respond to steroids (JAMA 2002;288:862–71). In that trial, 300 patients were tested for response to ACTH, and all were treated with glucocorticoid and mineralocorticoid supplementation. At 28 days, survival was higher in those patients who had tested positive for adrenal insufficiency; no benefit was seen in the other patients.

A closer examination of that study indicated that 24% of the patients were given etomidate, which is an adrenal suppressant and is no longer the standard of care. Also, subjects had a higher average dose of pressors at enrollment, an indication that they were sicker than the patients in the CORTICUS trial, and they received fludrocortisone rather than hydrocortisone, Dr. Schulman said.

The results of three trials have called into question the role of drotrecogin alfa (recombinant human activated protein C) in the management of sepsis. Both the RESOLVE (Resolution of Organ Failure in Pediatric Patients With Severe Sepsis) trial in children and the ADDRESS (Administration of Drotrecogin Alfa in Early Severe Sepsis) trial in low-risk patients randomized to recombinant human activated protein C or placebo were stopped early in light of their small chance of benefit and an increased risk of serious bleeding in the ADDRESS trial.

In the third ongoing trial, called PROWESS (Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis), patients with septic shock or severe sepsis were randomized to a continuous infusion of 24 mcg/kg per hour of recombinant human activated protein C for 96 hours. During the course of the trial, however, the protocol was changed to severe sepsis only, the manufacturing process of the product was changed, and the placebo was changed from normal saline to 0.1% albumin. Before the protocol change, no difference was noted in 28-day mortality. After the change, a 0.71 relative risk was noted in the treatment group, Dr. Schulman said.

SAN DIEGO — Recent trial results have quelled much of the excitement about the promise of interventions designed to reduce the risk of death resulting from sepsis. Yet within the mostly negative results of those studies, there are some findings that help to define appropriate treatment approaches, according to Dr. David A. Schulman, chief of pulmonary and critical care medicine at Emory University Hospital, Atlanta.

The findings of VASST (Vasopressin and Septic Shock Trial) found no difference in 28-day mortality among 778 patients randomized to norepinephrine or to norepinephrine plus vasopressin. Although the difference was not statistically significant, those who had less severe shock and were started on less than 15 mcg/min of norepinephrine had a lower mortality rate with the addition of vasopressin (26.5%), compared with the rate seen in those given norepinephrine alone (35.7%). The result persisted at 90 days (N. Engl. J. Med. 2008;358:877–87).

Thus, vasopressin can be used as a second agent in the hypotensive, septic patient who is on a moderate dose of another pressor, Dr. Schulman said in a presentation at the annual meeting of the Society of Hospital Medicine. Vasopressin, dosed at 0.03 U/min, may be added to norepinephrine with an anticipated effect equivalent to that of norepinephrine alone. Vasopressin should not be titrated, and it should be used only as a second agent. Vasopressin should be used with caution in patients with significant myocardial dysfunction (cardiac index less than 2.1 L/min per m

The recent CORTICUS (Corticosteroid Therapy of Septic Shock) trial concluded that steroids are no better than placebo for reducing mortality, yet the role of steroid therapy in the treatment of septic shock is “more unclear than ever,” Dr. Schulman said. Current data do not provide sufficient evidence that low doses of corticosteroids are harmful in septic patients without relative adrenal insufficiency. The current advisory is to “implement a combination of hydrocortisone and fludrocortisone if a patient remains hypotensive with presumed septic shock for [more than 1 hour] after administration of appropriate fluids and pressors.”

CORTICUS, a randomized controlled trial of 500 septic shock patients, found that 28-day mortality was comparable whether subjects were given placebo or hydrocortisone at 50 mg every 6 hours for 5 days followed by a 6-day taper. Furthermore, the trial discerned no difference in response to steroids among patients with relative adrenal insufficiency. Shock resolved faster in the steroid-treated subjects, but they didn't live any longer than patients who were not given steroids. Also, steroids were associated with an increased risk of hyperglycemia, recurrent sepsis, and recurrent shock (N. Engl. J. Med. 2008;358:111–24).

The findings countered those of the practice-changing study that held that adrenal insufficiency predicted which vasopressor-refractory sepsis patients would respond to steroids (JAMA 2002;288:862–71). In that trial, 300 patients were tested for response to ACTH, and all were treated with glucocorticoid and mineralocorticoid supplementation. At 28 days, survival was higher in those patients who had tested positive for adrenal insufficiency; no benefit was seen in the other patients.

A closer examination of that study indicated that 24% of the patients were given etomidate, which is an adrenal suppressant and is no longer the standard of care. Also, subjects had a higher average dose of pressors at enrollment, an indication that they were sicker than the patients in the CORTICUS trial, and they received fludrocortisone rather than hydrocortisone, Dr. Schulman said.

The results of three trials have called into question the role of drotrecogin alfa (recombinant human activated protein C) in the management of sepsis. Both the RESOLVE (Resolution of Organ Failure in Pediatric Patients With Severe Sepsis) trial in children and the ADDRESS (Administration of Drotrecogin Alfa in Early Severe Sepsis) trial in low-risk patients randomized to recombinant human activated protein C or placebo were stopped early in light of their small chance of benefit and an increased risk of serious bleeding in the ADDRESS trial.

In the third ongoing trial, called PROWESS (Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis), patients with septic shock or severe sepsis were randomized to a continuous infusion of 24 mcg/kg per hour of recombinant human activated protein C for 96 hours. During the course of the trial, however, the protocol was changed to severe sepsis only, the manufacturing process of the product was changed, and the placebo was changed from normal saline to 0.1% albumin. Before the protocol change, no difference was noted in 28-day mortality. After the change, a 0.71 relative risk was noted in the treatment group, Dr. Schulman said.

SAN DIEGO — Recent trial results have quelled much of the excitement about the promise of interventions designed to reduce the risk of death resulting from sepsis. Yet within the mostly negative results of those studies, there are some findings that help to define appropriate treatment approaches, according to Dr. David A. Schulman, chief of pulmonary and critical care medicine at Emory University Hospital, Atlanta.

The findings of VASST (Vasopressin and Septic Shock Trial) found no difference in 28-day mortality among 778 patients randomized to norepinephrine or to norepinephrine plus vasopressin. Although the difference was not statistically significant, those who had less severe shock and were started on less than 15 mcg/min of norepinephrine had a lower mortality rate with the addition of vasopressin (26.5%), compared with the rate seen in those given norepinephrine alone (35.7%). The result persisted at 90 days (N. Engl. J. Med. 2008;358:877–87).

Thus, vasopressin can be used as a second agent in the hypotensive, septic patient who is on a moderate dose of another pressor, Dr. Schulman said in a presentation at the annual meeting of the Society of Hospital Medicine. Vasopressin, dosed at 0.03 U/min, may be added to norepinephrine with an anticipated effect equivalent to that of norepinephrine alone. Vasopressin should not be titrated, and it should be used only as a second agent. Vasopressin should be used with caution in patients with significant myocardial dysfunction (cardiac index less than 2.1 L/min per m

The recent CORTICUS (Corticosteroid Therapy of Septic Shock) trial concluded that steroids are no better than placebo for reducing mortality, yet the role of steroid therapy in the treatment of septic shock is “more unclear than ever,” Dr. Schulman said. Current data do not provide sufficient evidence that low doses of corticosteroids are harmful in septic patients without relative adrenal insufficiency. The current advisory is to “implement a combination of hydrocortisone and fludrocortisone if a patient remains hypotensive with presumed septic shock for [more than 1 hour] after administration of appropriate fluids and pressors.”

CORTICUS, a randomized controlled trial of 500 septic shock patients, found that 28-day mortality was comparable whether subjects were given placebo or hydrocortisone at 50 mg every 6 hours for 5 days followed by a 6-day taper. Furthermore, the trial discerned no difference in response to steroids among patients with relative adrenal insufficiency. Shock resolved faster in the steroid-treated subjects, but they didn't live any longer than patients who were not given steroids. Also, steroids were associated with an increased risk of hyperglycemia, recurrent sepsis, and recurrent shock (N. Engl. J. Med. 2008;358:111–24).

The findings countered those of the practice-changing study that held that adrenal insufficiency predicted which vasopressor-refractory sepsis patients would respond to steroids (JAMA 2002;288:862–71). In that trial, 300 patients were tested for response to ACTH, and all were treated with glucocorticoid and mineralocorticoid supplementation. At 28 days, survival was higher in those patients who had tested positive for adrenal insufficiency; no benefit was seen in the other patients.

A closer examination of that study indicated that 24% of the patients were given etomidate, which is an adrenal suppressant and is no longer the standard of care. Also, subjects had a higher average dose of pressors at enrollment, an indication that they were sicker than the patients in the CORTICUS trial, and they received fludrocortisone rather than hydrocortisone, Dr. Schulman said.

The results of three trials have called into question the role of drotrecogin alfa (recombinant human activated protein C) in the management of sepsis. Both the RESOLVE (Resolution of Organ Failure in Pediatric Patients With Severe Sepsis) trial in children and the ADDRESS (Administration of Drotrecogin Alfa in Early Severe Sepsis) trial in low-risk patients randomized to recombinant human activated protein C or placebo were stopped early in light of their small chance of benefit and an increased risk of serious bleeding in the ADDRESS trial.

In the third ongoing trial, called PROWESS (Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis), patients with septic shock or severe sepsis were randomized to a continuous infusion of 24 mcg/kg per hour of recombinant human activated protein C for 96 hours. During the course of the trial, however, the protocol was changed to severe sepsis only, the manufacturing process of the product was changed, and the placebo was changed from normal saline to 0.1% albumin. Before the protocol change, no difference was noted in 28-day mortality. After the change, a 0.71 relative risk was noted in the treatment group, Dr. Schulman said.

CHICAGO – Vitamin E supplementation at doses of 2,000 IU/day appeared to be associated with improved survival in a retrospective case analysis of patients who had Alzheimer's disease.

The results, presented at the annual meeting of the American Academy of Neurology, were seen in a retrospective case analysis of 847 patients seen between 1990 and 2004 at the Alzheimer's Disease and Memory Disorders Center at Baylor College of Medicine, Houston.

The results do not indicate that high-dose vitamin E was associated with an increased risk of death in Alzheimer's patients, as has been seen in large studies of vitamin E for prevention of cardiovascular events.

All of the patients studied had probable or mixed Alzheimer's disease. About two-thirds of the subjects took 2,000 IU of vitamin E, with or without a cholinesterase inhibitor; less than 10% took vitamin E alone; and approximately 15% did not take any antidementia drug.

For those taking vitamin E, with or without a cholinesterase inhibitor, there was a 26% reduction in risk of dying (statistically significant at P = .009) at any time interval of the analysis, compared with those not taking vitamin E.

The prescribing of high-dose vitamin E in Alzheimer's disease gained popularity after a 1997 study indicated that vitamin E at doses of 2,000 IU/day appeared to slow the disease's progression. That approach fell into disfavor, however, when a meta-analysis of 19 randomized controlled trials involving more than 135,000 participants found that vitamin E supplementation for at least 1 year at doses greater than 400 IU/day was associated with increased all-cause mortality (Ann. Intern. Med. 2005;142:37–46).

Valory Pavlik, Ph.D., one of the investigators in the Baylor study, said that many of the studies in the meta-analysis examined vitamin E for prevention of cardiovascular events. Vitamin E was not being used for treatment as it was in the Baylor patients with Alzheimer's disease.

To determine whether the risk of death was greater for Alzheimer's patients, Dr. Pavlik and her associates undertook their analysis based on patients who were taking vitamin E during the time before high-dose vitamin E fell into disfavor.

In the Baylor study, patients averaged 74 years old and ranged in age from 65 to 83 years at their first visit to the Baylor center. Two-thirds of the patients were women; they were followed up for a median time of 5 years, with a range of 1–15 years.

Neuropsychological scores, clinical assessments, and medication histories were collected using standardized protocols. Vital status was ascertained through contact with family members or death index searches. Time-dependent Cox proportional hazards modeling was used to calculate all-cause mortality hazard ratios for vitamin E alone, or in combination with a cholinesterase inhibitor, adjusting for demographics, duration of symptoms at diagnosis, and baseline disease severity.

The adjusted hazard ratio associated with vitamin E (with or without a cholinesterase inhibitor) was 0.74 (95% CI = 0.59–0.92,P = .008), and for cholinesterase inhibitor use (with or without vitamin E), was 0.91 (95% CI = 0.72–1.14, P = .393).

The hazard ratios corresponding to mutually exclusive treatment categories (reference group = no drug treatment) were 0.79 (P = .069) for vitamin E with another drug, 1.1 (P = .515) for cholinesterase inhibitor use only, and 0.82 (P = .341) for vitamin E only.

Dr. Pavlik emphasized that this was an observational study and not a randomized controlled trial.

Alternatively, this group of patients was followed for longer times than were many of the patients in clinical trials. Further, survival rates were comparable or better among patients on cholinesterase inhibitors who also took vitamin E supplements at doses of 2,000 IU/day than they were among patients on cholinesterase inhibitors alone.

During a press conference, Dr. Pavlik said the lowest effective dose for vitamin E in Alzheimer's disease has not been determined. A dose-response study conducted at Vanderbilt University, Nashville, Tenn., has indicated that the antioxidant action of vitamin E begins to occur at doses exceeding 1,000 IU/day.

CHICAGO – Vitamin E supplementation at doses of 2,000 IU/day appeared to be associated with improved survival in a retrospective case analysis of patients who had Alzheimer's disease.

The results, presented at the annual meeting of the American Academy of Neurology, were seen in a retrospective case analysis of 847 patients seen between 1990 and 2004 at the Alzheimer's Disease and Memory Disorders Center at Baylor College of Medicine, Houston.

The results do not indicate that high-dose vitamin E was associated with an increased risk of death in Alzheimer's patients, as has been seen in large studies of vitamin E for prevention of cardiovascular events.

All of the patients studied had probable or mixed Alzheimer's disease. About two-thirds of the subjects took 2,000 IU of vitamin E, with or without a cholinesterase inhibitor; less than 10% took vitamin E alone; and approximately 15% did not take any antidementia drug.

For those taking vitamin E, with or without a cholinesterase inhibitor, there was a 26% reduction in risk of dying (statistically significant at P = .009) at any time interval of the analysis, compared with those not taking vitamin E.

The prescribing of high-dose vitamin E in Alzheimer's disease gained popularity after a 1997 study indicated that vitamin E at doses of 2,000 IU/day appeared to slow the disease's progression. That approach fell into disfavor, however, when a meta-analysis of 19 randomized controlled trials involving more than 135,000 participants found that vitamin E supplementation for at least 1 year at doses greater than 400 IU/day was associated with increased all-cause mortality (Ann. Intern. Med. 2005;142:37–46).

Valory Pavlik, Ph.D., one of the investigators in the Baylor study, said that many of the studies in the meta-analysis examined vitamin E for prevention of cardiovascular events. Vitamin E was not being used for treatment as it was in the Baylor patients with Alzheimer's disease.

To determine whether the risk of death was greater for Alzheimer's patients, Dr. Pavlik and her associates undertook their analysis based on patients who were taking vitamin E during the time before high-dose vitamin E fell into disfavor.

In the Baylor study, patients averaged 74 years old and ranged in age from 65 to 83 years at their first visit to the Baylor center. Two-thirds of the patients were women; they were followed up for a median time of 5 years, with a range of 1–15 years.

Neuropsychological scores, clinical assessments, and medication histories were collected using standardized protocols. Vital status was ascertained through contact with family members or death index searches. Time-dependent Cox proportional hazards modeling was used to calculate all-cause mortality hazard ratios for vitamin E alone, or in combination with a cholinesterase inhibitor, adjusting for demographics, duration of symptoms at diagnosis, and baseline disease severity.

The adjusted hazard ratio associated with vitamin E (with or without a cholinesterase inhibitor) was 0.74 (95% CI = 0.59–0.92,P = .008), and for cholinesterase inhibitor use (with or without vitamin E), was 0.91 (95% CI = 0.72–1.14, P = .393).

The hazard ratios corresponding to mutually exclusive treatment categories (reference group = no drug treatment) were 0.79 (P = .069) for vitamin E with another drug, 1.1 (P = .515) for cholinesterase inhibitor use only, and 0.82 (P = .341) for vitamin E only.

Dr. Pavlik emphasized that this was an observational study and not a randomized controlled trial.

Alternatively, this group of patients was followed for longer times than were many of the patients in clinical trials. Further, survival rates were comparable or better among patients on cholinesterase inhibitors who also took vitamin E supplements at doses of 2,000 IU/day than they were among patients on cholinesterase inhibitors alone.

During a press conference, Dr. Pavlik said the lowest effective dose for vitamin E in Alzheimer's disease has not been determined. A dose-response study conducted at Vanderbilt University, Nashville, Tenn., has indicated that the antioxidant action of vitamin E begins to occur at doses exceeding 1,000 IU/day.

CHICAGO – Vitamin E supplementation at doses of 2,000 IU/day appeared to be associated with improved survival in a retrospective case analysis of patients who had Alzheimer's disease.

The results, presented at the annual meeting of the American Academy of Neurology, were seen in a retrospective case analysis of 847 patients seen between 1990 and 2004 at the Alzheimer's Disease and Memory Disorders Center at Baylor College of Medicine, Houston.

The results do not indicate that high-dose vitamin E was associated with an increased risk of death in Alzheimer's patients, as has been seen in large studies of vitamin E for prevention of cardiovascular events.

All of the patients studied had probable or mixed Alzheimer's disease. About two-thirds of the subjects took 2,000 IU of vitamin E, with or without a cholinesterase inhibitor; less than 10% took vitamin E alone; and approximately 15% did not take any antidementia drug.

For those taking vitamin E, with or without a cholinesterase inhibitor, there was a 26% reduction in risk of dying (statistically significant at P = .009) at any time interval of the analysis, compared with those not taking vitamin E.

The prescribing of high-dose vitamin E in Alzheimer's disease gained popularity after a 1997 study indicated that vitamin E at doses of 2,000 IU/day appeared to slow the disease's progression. That approach fell into disfavor, however, when a meta-analysis of 19 randomized controlled trials involving more than 135,000 participants found that vitamin E supplementation for at least 1 year at doses greater than 400 IU/day was associated with increased all-cause mortality (Ann. Intern. Med. 2005;142:37–46).

Valory Pavlik, Ph.D., one of the investigators in the Baylor study, said that many of the studies in the meta-analysis examined vitamin E for prevention of cardiovascular events. Vitamin E was not being used for treatment as it was in the Baylor patients with Alzheimer's disease.

To determine whether the risk of death was greater for Alzheimer's patients, Dr. Pavlik and her associates undertook their analysis based on patients who were taking vitamin E during the time before high-dose vitamin E fell into disfavor.

In the Baylor study, patients averaged 74 years old and ranged in age from 65 to 83 years at their first visit to the Baylor center. Two-thirds of the patients were women; they were followed up for a median time of 5 years, with a range of 1–15 years.

Neuropsychological scores, clinical assessments, and medication histories were collected using standardized protocols. Vital status was ascertained through contact with family members or death index searches. Time-dependent Cox proportional hazards modeling was used to calculate all-cause mortality hazard ratios for vitamin E alone, or in combination with a cholinesterase inhibitor, adjusting for demographics, duration of symptoms at diagnosis, and baseline disease severity.

The adjusted hazard ratio associated with vitamin E (with or without a cholinesterase inhibitor) was 0.74 (95% CI = 0.59–0.92,P = .008), and for cholinesterase inhibitor use (with or without vitamin E), was 0.91 (95% CI = 0.72–1.14, P = .393).

The hazard ratios corresponding to mutually exclusive treatment categories (reference group = no drug treatment) were 0.79 (P = .069) for vitamin E with another drug, 1.1 (P = .515) for cholinesterase inhibitor use only, and 0.82 (P = .341) for vitamin E only.

Dr. Pavlik emphasized that this was an observational study and not a randomized controlled trial.

Alternatively, this group of patients was followed for longer times than were many of the patients in clinical trials. Further, survival rates were comparable or better among patients on cholinesterase inhibitors who also took vitamin E supplements at doses of 2,000 IU/day than they were among patients on cholinesterase inhibitors alone.

During a press conference, Dr. Pavlik said the lowest effective dose for vitamin E in Alzheimer's disease has not been determined. A dose-response study conducted at Vanderbilt University, Nashville, Tenn., has indicated that the antioxidant action of vitamin E begins to occur at doses exceeding 1,000 IU/day.

CHICAGO – Men have more mild cognitive impairment than women do, yet there is no gender difference in the prevalence of dementia, according to the results of one of the first studies to measure mild cognitive impairment prospectively in a population-based setting.

The findings, reported by Dr. Rosebud O. Roberts at the annual meeting of the American Academy of Neurology, suggest that dementia progresses either faster in women or slower in men.

For the ongoing study, called the Mayo Clinic Study of Aging, mild cognitive impairment was evaluated in a population sample from Olmstead County, Minn. The sampling scheme aimed for equal numbers of individuals in each gender and age group. The 70- to 79-year-old group included 490 women and 596 men and the 80- to 89-year-old group included 512 women and 452 men. For both age groups, there were 1,002 women and 1,048 men.

Either a nurse, physician, or neuropsychologist evaluated each individual using face-to-face measures. Subjects were evaluated in four domains–memory, executive function, language, and visual/spatial skills. Mild cognitive impairment (MCI) was defined as impairment in one or more domains or an overall mild decline across cognitive abilities that is greater than would be expected for an individual's age or education but is insufficient to interfere with social and occupational functioning.

Based on these evaluations, 74% of the group had normal cognition, 16% had mild cognitive impairment, and 10% had dementia. Of the nearly 2,000 study participants without dementia, 51% were male, 47% had less than 12 years of education, 52% were 80–89 years old, and 61% were married.

Subjects were studied prospectively beginning in October 2004 and follow-ups will continue through 2010. This differs from most other studies of MCI, which had the limitations of applying MCI criteria to previously collected data or were conducted in study samples, such as those attending memory clinics, who might not be representative of the general population, said Dr. Roberts, an epidemiologist at the Mayo Clinic, Rochester, Minn.

In men, the prevalence of mild cognitive impairment steadily increased from about 10% at age 70 and suddenly spiked after age 85 to affect 40%. In women, the rate rose more slowly and the prevalence was far lower, peaking at less than 20% at age 85.

Even after the data were corrected for age plus education, marital status, and disease burden, women had less cognitive impairment but comparable rates of dementia, compared with men, Dr. Roberts said.

“We found the overall prevalence of mild cognitive impairment is quite high–over 16%,” said Dr. Roberts. “But perhaps the more surprising finding is the higher prevalence of MCI in men with the comparable prevalence of dementia for men and women.” Several possible explanations for this disparity include a prevalence of risk factors in middle age vs. later life, the progression rate from MCI to dementia, and death among persons with MCI.

Dr. Roberts said that she and her coinvestigators are adding on another 1,000 study participants to continue the follow-up study and are applying for additional funding. The study was supported by grants from the National Institutes of Health and the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer's Disease Research Program.

In men, the prevalence of mild cognitive impairment spiked after age 85 to affect 40%. DR. ROBERTS

CHICAGO – Men have more mild cognitive impairment than women do, yet there is no gender difference in the prevalence of dementia, according to the results of one of the first studies to measure mild cognitive impairment prospectively in a population-based setting.

The findings, reported by Dr. Rosebud O. Roberts at the annual meeting of the American Academy of Neurology, suggest that dementia progresses either faster in women or slower in men.

For the ongoing study, called the Mayo Clinic Study of Aging, mild cognitive impairment was evaluated in a population sample from Olmstead County, Minn. The sampling scheme aimed for equal numbers of individuals in each gender and age group. The 70- to 79-year-old group included 490 women and 596 men and the 80- to 89-year-old group included 512 women and 452 men. For both age groups, there were 1,002 women and 1,048 men.

Either a nurse, physician, or neuropsychologist evaluated each individual using face-to-face measures. Subjects were evaluated in four domains–memory, executive function, language, and visual/spatial skills. Mild cognitive impairment (MCI) was defined as impairment in one or more domains or an overall mild decline across cognitive abilities that is greater than would be expected for an individual's age or education but is insufficient to interfere with social and occupational functioning.

Based on these evaluations, 74% of the group had normal cognition, 16% had mild cognitive impairment, and 10% had dementia. Of the nearly 2,000 study participants without dementia, 51% were male, 47% had less than 12 years of education, 52% were 80–89 years old, and 61% were married.

Subjects were studied prospectively beginning in October 2004 and follow-ups will continue through 2010. This differs from most other studies of MCI, which had the limitations of applying MCI criteria to previously collected data or were conducted in study samples, such as those attending memory clinics, who might not be representative of the general population, said Dr. Roberts, an epidemiologist at the Mayo Clinic, Rochester, Minn.

In men, the prevalence of mild cognitive impairment steadily increased from about 10% at age 70 and suddenly spiked after age 85 to affect 40%. In women, the rate rose more slowly and the prevalence was far lower, peaking at less than 20% at age 85.

Even after the data were corrected for age plus education, marital status, and disease burden, women had less cognitive impairment but comparable rates of dementia, compared with men, Dr. Roberts said.

“We found the overall prevalence of mild cognitive impairment is quite high–over 16%,” said Dr. Roberts. “But perhaps the more surprising finding is the higher prevalence of MCI in men with the comparable prevalence of dementia for men and women.” Several possible explanations for this disparity include a prevalence of risk factors in middle age vs. later life, the progression rate from MCI to dementia, and death among persons with MCI.

Dr. Roberts said that she and her coinvestigators are adding on another 1,000 study participants to continue the follow-up study and are applying for additional funding. The study was supported by grants from the National Institutes of Health and the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer's Disease Research Program.

In men, the prevalence of mild cognitive impairment spiked after age 85 to affect 40%. DR. ROBERTS

CHICAGO – Men have more mild cognitive impairment than women do, yet there is no gender difference in the prevalence of dementia, according to the results of one of the first studies to measure mild cognitive impairment prospectively in a population-based setting.

The findings, reported by Dr. Rosebud O. Roberts at the annual meeting of the American Academy of Neurology, suggest that dementia progresses either faster in women or slower in men.

For the ongoing study, called the Mayo Clinic Study of Aging, mild cognitive impairment was evaluated in a population sample from Olmstead County, Minn. The sampling scheme aimed for equal numbers of individuals in each gender and age group. The 70- to 79-year-old group included 490 women and 596 men and the 80- to 89-year-old group included 512 women and 452 men. For both age groups, there were 1,002 women and 1,048 men.

Either a nurse, physician, or neuropsychologist evaluated each individual using face-to-face measures. Subjects were evaluated in four domains–memory, executive function, language, and visual/spatial skills. Mild cognitive impairment (MCI) was defined as impairment in one or more domains or an overall mild decline across cognitive abilities that is greater than would be expected for an individual's age or education but is insufficient to interfere with social and occupational functioning.

Based on these evaluations, 74% of the group had normal cognition, 16% had mild cognitive impairment, and 10% had dementia. Of the nearly 2,000 study participants without dementia, 51% were male, 47% had less than 12 years of education, 52% were 80–89 years old, and 61% were married.

Subjects were studied prospectively beginning in October 2004 and follow-ups will continue through 2010. This differs from most other studies of MCI, which had the limitations of applying MCI criteria to previously collected data or were conducted in study samples, such as those attending memory clinics, who might not be representative of the general population, said Dr. Roberts, an epidemiologist at the Mayo Clinic, Rochester, Minn.

In men, the prevalence of mild cognitive impairment steadily increased from about 10% at age 70 and suddenly spiked after age 85 to affect 40%. In women, the rate rose more slowly and the prevalence was far lower, peaking at less than 20% at age 85.

Even after the data were corrected for age plus education, marital status, and disease burden, women had less cognitive impairment but comparable rates of dementia, compared with men, Dr. Roberts said.

“We found the overall prevalence of mild cognitive impairment is quite high–over 16%,” said Dr. Roberts. “But perhaps the more surprising finding is the higher prevalence of MCI in men with the comparable prevalence of dementia for men and women.” Several possible explanations for this disparity include a prevalence of risk factors in middle age vs. later life, the progression rate from MCI to dementia, and death among persons with MCI.

Dr. Roberts said that she and her coinvestigators are adding on another 1,000 study participants to continue the follow-up study and are applying for additional funding. The study was supported by grants from the National Institutes of Health and the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer's Disease Research Program.

In men, the prevalence of mild cognitive impairment spiked after age 85 to affect 40%. DR. ROBERTS

SAN DIEGO — Noninvasive positive pressure ventilation should “absolutely” be used for nearly every patient hospitalized for an acute exacerbation of chronic obstructive pulmonary disease, said Dr. Daniel D. Dressler, director of education for the section of hospital medicine at Emory University, Atlanta.

“Learn how to use this therapy,” Dr. Dressler advised at the annual meeting of the Society of Hospital Medicine. Although contraindicated in COPD patients who have malignant arrhythmias, refractory hypoxemia, cardiac or respiratory arrest, or hemodynamic instability, adding noninvasive positive pressure ventilation (NPPV) to usual care has been shown in a Cochrane meta-analysis of 14 clinical trials to save one life for every 10 COPD inpatients treated and to avoid intubation for 1 of every 4 patients treated, compared with usual care alone. Further, NPPV has been shown to lower the average length of stay by 3.2 days.

Dr. Dressler advised starting NPPV in the emergency department if possible, since “early intervention likely improves outcomes.” Monitor these patients closely with arterial blood gases at 30–60 minutes after initiating or changing NPPV settings. Don't wait 2 hours before taking these measures.

The usual starting pressure level is 10/5 cm of water pressure, and the settings are titrated upward to tolerance with the assistance of the respiratory therapist, he said.

Based on the Cochrane review, those who respond best to NPPV are patients with pH values less than 7.3. But that finding largely may reflect the level of care at individual hospitals rather than the likelihood that the therapy was poorly tolerated by individual patients. In a recent study of COPD patients with pH values of 7.35 and higher, NPPV did not reduce the risk of death or intubation, but it did speed up reduction in the partial pressure of carbon dioxide (pCO₂) and reduced average length of stay from 10.2 days to 5.5 days (Eur. J. Intern. Med. 2007;18:524–30).

Antibiotic treatment for nearly all hospitalized patients with COPD exacerbations also reduces length of stay and is strongly supported by evidence, according to Cochrane analysis. On average, treating eight patients will prevent one death, and treating three patients will prevent one treatment failure.

The evidence for the choice of antibiotic is not great, however, and not all hospitalized patients with COPD exacerbations are likely to need antibiotic therapy. Among the tests investigators are examining, as tools to guide antibiotic prescribing, are measures of procalcitonin levels, with values less than 0.1 ng/mL discouraging the use of antibiotics and values exceeding 0.25 ng/mL encouraging the use of antibiotics. When compared with patients treated empirically by their physicians, outcomes did not differ for those treated based on procalcitonin levels, but using the test did have the desired result of reducing antibiotic use. One had to test four patients in order to avoid prescribing antibiotics for one patient. The procalcitonin test is not broadly available and is estimated to cost about $250, but Dr. Dressler predicted more use of this test in the next 2–4 years.

The evidence is similarly strong for prescribing systemic steroids and inhaled bronchodilators for these patients. Oxygen obviously is prescribed as well, despite a lack of studies showing it to be appropriate for COPD exacerbations among inpatients.

During hospitalization and at discharge of COPD patients, Dr. Dressler advised tobacco cessation counseling and, if not previously administered, a pneumonia vaccine and an annual influenza vaccine. Prophylaxis for venous thromboembolism is standard during the hospital stay.

Patients' home medication regimens should be augmented with a long-acting β-agonist and corticosteroid inhaler. Single agents have been shown to reduce the risk of exacerbations, but they have not been shown to reduce mortality. Combination therapy has been shown to reduce exacerbations and avoids one death for every 53 patients treated.

SAN DIEGO — Noninvasive positive pressure ventilation should “absolutely” be used for nearly every patient hospitalized for an acute exacerbation of chronic obstructive pulmonary disease, said Dr. Daniel D. Dressler, director of education for the section of hospital medicine at Emory University, Atlanta.

“Learn how to use this therapy,” Dr. Dressler advised at the annual meeting of the Society of Hospital Medicine. Although contraindicated in COPD patients who have malignant arrhythmias, refractory hypoxemia, cardiac or respiratory arrest, or hemodynamic instability, adding noninvasive positive pressure ventilation (NPPV) to usual care has been shown in a Cochrane meta-analysis of 14 clinical trials to save one life for every 10 COPD inpatients treated and to avoid intubation for 1 of every 4 patients treated, compared with usual care alone. Further, NPPV has been shown to lower the average length of stay by 3.2 days.

Dr. Dressler advised starting NPPV in the emergency department if possible, since “early intervention likely improves outcomes.” Monitor these patients closely with arterial blood gases at 30–60 minutes after initiating or changing NPPV settings. Don't wait 2 hours before taking these measures.

The usual starting pressure level is 10/5 cm of water pressure, and the settings are titrated upward to tolerance with the assistance of the respiratory therapist, he said.

Based on the Cochrane review, those who respond best to NPPV are patients with pH values less than 7.3. But that finding largely may reflect the level of care at individual hospitals rather than the likelihood that the therapy was poorly tolerated by individual patients. In a recent study of COPD patients with pH values of 7.35 and higher, NPPV did not reduce the risk of death or intubation, but it did speed up reduction in the partial pressure of carbon dioxide (pCO₂) and reduced average length of stay from 10.2 days to 5.5 days (Eur. J. Intern. Med. 2007;18:524–30).

Antibiotic treatment for nearly all hospitalized patients with COPD exacerbations also reduces length of stay and is strongly supported by evidence, according to Cochrane analysis. On average, treating eight patients will prevent one death, and treating three patients will prevent one treatment failure.

The evidence for the choice of antibiotic is not great, however, and not all hospitalized patients with COPD exacerbations are likely to need antibiotic therapy. Among the tests investigators are examining, as tools to guide antibiotic prescribing, are measures of procalcitonin levels, with values less than 0.1 ng/mL discouraging the use of antibiotics and values exceeding 0.25 ng/mL encouraging the use of antibiotics. When compared with patients treated empirically by their physicians, outcomes did not differ for those treated based on procalcitonin levels, but using the test did have the desired result of reducing antibiotic use. One had to test four patients in order to avoid prescribing antibiotics for one patient. The procalcitonin test is not broadly available and is estimated to cost about $250, but Dr. Dressler predicted more use of this test in the next 2–4 years.

The evidence is similarly strong for prescribing systemic steroids and inhaled bronchodilators for these patients. Oxygen obviously is prescribed as well, despite a lack of studies showing it to be appropriate for COPD exacerbations among inpatients.

During hospitalization and at discharge of COPD patients, Dr. Dressler advised tobacco cessation counseling and, if not previously administered, a pneumonia vaccine and an annual influenza vaccine. Prophylaxis for venous thromboembolism is standard during the hospital stay.

Patients' home medication regimens should be augmented with a long-acting β-agonist and corticosteroid inhaler. Single agents have been shown to reduce the risk of exacerbations, but they have not been shown to reduce mortality. Combination therapy has been shown to reduce exacerbations and avoids one death for every 53 patients treated.

SAN DIEGO — Noninvasive positive pressure ventilation should “absolutely” be used for nearly every patient hospitalized for an acute exacerbation of chronic obstructive pulmonary disease, said Dr. Daniel D. Dressler, director of education for the section of hospital medicine at Emory University, Atlanta.

“Learn how to use this therapy,” Dr. Dressler advised at the annual meeting of the Society of Hospital Medicine. Although contraindicated in COPD patients who have malignant arrhythmias, refractory hypoxemia, cardiac or respiratory arrest, or hemodynamic instability, adding noninvasive positive pressure ventilation (NPPV) to usual care has been shown in a Cochrane meta-analysis of 14 clinical trials to save one life for every 10 COPD inpatients treated and to avoid intubation for 1 of every 4 patients treated, compared with usual care alone. Further, NPPV has been shown to lower the average length of stay by 3.2 days.

Dr. Dressler advised starting NPPV in the emergency department if possible, since “early intervention likely improves outcomes.” Monitor these patients closely with arterial blood gases at 30–60 minutes after initiating or changing NPPV settings. Don't wait 2 hours before taking these measures.

The usual starting pressure level is 10/5 cm of water pressure, and the settings are titrated upward to tolerance with the assistance of the respiratory therapist, he said.

Based on the Cochrane review, those who respond best to NPPV are patients with pH values less than 7.3. But that finding largely may reflect the level of care at individual hospitals rather than the likelihood that the therapy was poorly tolerated by individual patients. In a recent study of COPD patients with pH values of 7.35 and higher, NPPV did not reduce the risk of death or intubation, but it did speed up reduction in the partial pressure of carbon dioxide (pCO₂) and reduced average length of stay from 10.2 days to 5.5 days (Eur. J. Intern. Med. 2007;18:524–30).

Antibiotic treatment for nearly all hospitalized patients with COPD exacerbations also reduces length of stay and is strongly supported by evidence, according to Cochrane analysis. On average, treating eight patients will prevent one death, and treating three patients will prevent one treatment failure.

The evidence for the choice of antibiotic is not great, however, and not all hospitalized patients with COPD exacerbations are likely to need antibiotic therapy. Among the tests investigators are examining, as tools to guide antibiotic prescribing, are measures of procalcitonin levels, with values less than 0.1 ng/mL discouraging the use of antibiotics and values exceeding 0.25 ng/mL encouraging the use of antibiotics. When compared with patients treated empirically by their physicians, outcomes did not differ for those treated based on procalcitonin levels, but using the test did have the desired result of reducing antibiotic use. One had to test four patients in order to avoid prescribing antibiotics for one patient. The procalcitonin test is not broadly available and is estimated to cost about $250, but Dr. Dressler predicted more use of this test in the next 2–4 years.

The evidence is similarly strong for prescribing systemic steroids and inhaled bronchodilators for these patients. Oxygen obviously is prescribed as well, despite a lack of studies showing it to be appropriate for COPD exacerbations among inpatients.

During hospitalization and at discharge of COPD patients, Dr. Dressler advised tobacco cessation counseling and, if not previously administered, a pneumonia vaccine and an annual influenza vaccine. Prophylaxis for venous thromboembolism is standard during the hospital stay.

Patients' home medication regimens should be augmented with a long-acting β-agonist and corticosteroid inhaler. Single agents have been shown to reduce the risk of exacerbations, but they have not been shown to reduce mortality. Combination therapy has been shown to reduce exacerbations and avoids one death for every 53 patients treated.

BOSTON – In soldiers returning from combat in Iraq, a self-reported history of migraine headaches was associated with at least twice the risk of symptoms of depression, posttraumatic stress disorder, and anxiety as was seen in similar soldiers without migraines, based on a study presented at the annual meeting of the American Academy of Neurology.

Migraines appear to be frequently associated with symptoms of psychiatric conditions in soldiers returning from deployment, Maj. Jay C. Erickson, MC, USA, said during a press conference at the meeting, where the study results were presented during a scientific poster session.

Although all soldiers returning from deployment undergo mental health screening, there is the possibility that headaches and symptoms of a psychiatric condition could present after such testing and possibly outside the Veterans Affairs health care system, he said. Mental health screening is warranted to assure that psychiatric disorders are identified and properly treated at that time.

Dr. Erickson of Madigan Army Medical Center at Fort Lewis in Tacoma, Wash., reported the findings from a health screening questionnaire completed by nearly 2,200 of 3,600 soldiers returning to Fort Lewis in Washington state after a 1-year combat duty deployment to Iraq. The questionnaires, completed within 90 days after the soldiers' return, indicated that nearly 20% had migraine headaches.

The study results are limited by their self-reported nature, which does not establish a diagnosis and is likely to result in more reports of symptoms; they also are limited by a lack of information about predeployment rates of migraine. Nevertheless, the rates are twice those seen in 20- to 40-year-olds in the general population and in men, who comprised 96% of the study participants.

Respondents answered 15 questions about the nature and frequency of any headaches in the last 3 months. They also completed the four-question Primary Care PTSD Screen (PC-PTSD) and the Patient Health Questionnaire (PHQ9) screen for depression and anxiety.

The responses indicated that 32% screened positive for depression, 22% for PTSD, and 9% for anxiety. Overall, 39% of respondents had at least one psychiatric condition. Respondents with migraines, compared with those without migraines, had much higher rates of depression symptoms (50% vs. 27%), symptoms of PTSD (39% vs. 18%), and anxiety symptoms (17% vs. 7%).

Migraine days per month were linked to a higher probability of a positive screen for depression and PTSD, but not a higher rate of anxiety symptoms. Those with migraine and depression symptoms had an average of 3.5 headaches days per month, compared with 2.5 days for those with migraine and no depression.

BOSTON – In soldiers returning from combat in Iraq, a self-reported history of migraine headaches was associated with at least twice the risk of symptoms of depression, posttraumatic stress disorder, and anxiety as was seen in similar soldiers without migraines, based on a study presented at the annual meeting of the American Academy of Neurology.

Migraines appear to be frequently associated with symptoms of psychiatric conditions in soldiers returning from deployment, Maj. Jay C. Erickson, MC, USA, said during a press conference at the meeting, where the study results were presented during a scientific poster session.

Although all soldiers returning from deployment undergo mental health screening, there is the possibility that headaches and symptoms of a psychiatric condition could present after such testing and possibly outside the Veterans Affairs health care system, he said. Mental health screening is warranted to assure that psychiatric disorders are identified and properly treated at that time.

Dr. Erickson of Madigan Army Medical Center at Fort Lewis in Tacoma, Wash., reported the findings from a health screening questionnaire completed by nearly 2,200 of 3,600 soldiers returning to Fort Lewis in Washington state after a 1-year combat duty deployment to Iraq. The questionnaires, completed within 90 days after the soldiers' return, indicated that nearly 20% had migraine headaches.

The study results are limited by their self-reported nature, which does not establish a diagnosis and is likely to result in more reports of symptoms; they also are limited by a lack of information about predeployment rates of migraine. Nevertheless, the rates are twice those seen in 20- to 40-year-olds in the general population and in men, who comprised 96% of the study participants.

Respondents answered 15 questions about the nature and frequency of any headaches in the last 3 months. They also completed the four-question Primary Care PTSD Screen (PC-PTSD) and the Patient Health Questionnaire (PHQ9) screen for depression and anxiety.

The responses indicated that 32% screened positive for depression, 22% for PTSD, and 9% for anxiety. Overall, 39% of respondents had at least one psychiatric condition. Respondents with migraines, compared with those without migraines, had much higher rates of depression symptoms (50% vs. 27%), symptoms of PTSD (39% vs. 18%), and anxiety symptoms (17% vs. 7%).

Migraine days per month were linked to a higher probability of a positive screen for depression and PTSD, but not a higher rate of anxiety symptoms. Those with migraine and depression symptoms had an average of 3.5 headaches days per month, compared with 2.5 days for those with migraine and no depression.

BOSTON – In soldiers returning from combat in Iraq, a self-reported history of migraine headaches was associated with at least twice the risk of symptoms of depression, posttraumatic stress disorder, and anxiety as was seen in similar soldiers without migraines, based on a study presented at the annual meeting of the American Academy of Neurology.

Migraines appear to be frequently associated with symptoms of psychiatric conditions in soldiers returning from deployment, Maj. Jay C. Erickson, MC, USA, said during a press conference at the meeting, where the study results were presented during a scientific poster session.

Although all soldiers returning from deployment undergo mental health screening, there is the possibility that headaches and symptoms of a psychiatric condition could present after such testing and possibly outside the Veterans Affairs health care system, he said. Mental health screening is warranted to assure that psychiatric disorders are identified and properly treated at that time.

Dr. Erickson of Madigan Army Medical Center at Fort Lewis in Tacoma, Wash., reported the findings from a health screening questionnaire completed by nearly 2,200 of 3,600 soldiers returning to Fort Lewis in Washington state after a 1-year combat duty deployment to Iraq. The questionnaires, completed within 90 days after the soldiers' return, indicated that nearly 20% had migraine headaches.

The study results are limited by their self-reported nature, which does not establish a diagnosis and is likely to result in more reports of symptoms; they also are limited by a lack of information about predeployment rates of migraine. Nevertheless, the rates are twice those seen in 20- to 40-year-olds in the general population and in men, who comprised 96% of the study participants.

Respondents answered 15 questions about the nature and frequency of any headaches in the last 3 months. They also completed the four-question Primary Care PTSD Screen (PC-PTSD) and the Patient Health Questionnaire (PHQ9) screen for depression and anxiety.

The responses indicated that 32% screened positive for depression, 22% for PTSD, and 9% for anxiety. Overall, 39% of respondents had at least one psychiatric condition. Respondents with migraines, compared with those without migraines, had much higher rates of depression symptoms (50% vs. 27%), symptoms of PTSD (39% vs. 18%), and anxiety symptoms (17% vs. 7%).

Migraine days per month were linked to a higher probability of a positive screen for depression and PTSD, but not a higher rate of anxiety symptoms. Those with migraine and depression symptoms had an average of 3.5 headaches days per month, compared with 2.5 days for those with migraine and no depression.

BOSTON – Sleep strengthens declarative memory, a finding that could one day be exploited to combat cognitive declines associated with dementia and neurologic disorders, Dr. Jeffrey M. Ellenbogen reported at the annual meeting of the American Academy of Neurology.

Sleep previously has been shown to improve motor, visual, and perceptual memories in humans, but this study is one of the first to substantiate a role for sleep in protecting declarative or consciously discussed memories from associative interference, said Dr. Ellenbogen of Harvard Medical School, Boston.

For the study, 48 healthy individuals, aged 18–30 years, were evaluated. No participants took any medications, and all had normal sleep patterns. Subjects were placed in one of four groups, and all were asked to remember 20 words, each of which was paired with an associated word cue.

All groups were tested for their abilities to remember the associated words 12 hours after learning the list. No one in the study was sleep deprived, but the subjects in two groups slept before testing and those in the other two groups did not.

The individuals who slept remembered 94% of the associations; those who did not sleep remembered 82% of the words. Those differences fell just short of statistical significance (P = .06).

The differences were marked, however, in the remaining two groups subjected to associative interference. For this part of the study, both groups also learned the same list of 20 word associations.

Again, one group slept and the other did not, and both groups were tested at 12 hours. But 12 minutes before they were tested on the list of words, both groups were subjected to associative interference. They were asked to learn a second list of 20 additional words related to the same word cues. Subjects who had slept and were introduced to the second list of words were able to recall 76% of the word associations from the first list. The subjects who had not slept were able to remember just 32% of the words. The difference in performance was highly significant (P less than .0001).

Sleep is ultimately a brain state that may hold the keys to understanding the neurobiology of memory consolidation, he said. The next step in his research will be to look at the relationships between sleep disorders, cognitive impairment, and neurologic disorders. Dr. Ellenbogen's research was supported by the National Institutes of Health and the University of Pennsylvania's Nassau Undergraduate Research Fund.

BOSTON – Sleep strengthens declarative memory, a finding that could one day be exploited to combat cognitive declines associated with dementia and neurologic disorders, Dr. Jeffrey M. Ellenbogen reported at the annual meeting of the American Academy of Neurology.

Sleep previously has been shown to improve motor, visual, and perceptual memories in humans, but this study is one of the first to substantiate a role for sleep in protecting declarative or consciously discussed memories from associative interference, said Dr. Ellenbogen of Harvard Medical School, Boston.

For the study, 48 healthy individuals, aged 18–30 years, were evaluated. No participants took any medications, and all had normal sleep patterns. Subjects were placed in one of four groups, and all were asked to remember 20 words, each of which was paired with an associated word cue.

All groups were tested for their abilities to remember the associated words 12 hours after learning the list. No one in the study was sleep deprived, but the subjects in two groups slept before testing and those in the other two groups did not.

The individuals who slept remembered 94% of the associations; those who did not sleep remembered 82% of the words. Those differences fell just short of statistical significance (P = .06).

The differences were marked, however, in the remaining two groups subjected to associative interference. For this part of the study, both groups also learned the same list of 20 word associations.

Again, one group slept and the other did not, and both groups were tested at 12 hours. But 12 minutes before they were tested on the list of words, both groups were subjected to associative interference. They were asked to learn a second list of 20 additional words related to the same word cues. Subjects who had slept and were introduced to the second list of words were able to recall 76% of the word associations from the first list. The subjects who had not slept were able to remember just 32% of the words. The difference in performance was highly significant (P less than .0001).

Sleep is ultimately a brain state that may hold the keys to understanding the neurobiology of memory consolidation, he said. The next step in his research will be to look at the relationships between sleep disorders, cognitive impairment, and neurologic disorders. Dr. Ellenbogen's research was supported by the National Institutes of Health and the University of Pennsylvania's Nassau Undergraduate Research Fund.

BOSTON – Sleep strengthens declarative memory, a finding that could one day be exploited to combat cognitive declines associated with dementia and neurologic disorders, Dr. Jeffrey M. Ellenbogen reported at the annual meeting of the American Academy of Neurology.

Sleep previously has been shown to improve motor, visual, and perceptual memories in humans, but this study is one of the first to substantiate a role for sleep in protecting declarative or consciously discussed memories from associative interference, said Dr. Ellenbogen of Harvard Medical School, Boston.

For the study, 48 healthy individuals, aged 18–30 years, were evaluated. No participants took any medications, and all had normal sleep patterns. Subjects were placed in one of four groups, and all were asked to remember 20 words, each of which was paired with an associated word cue.

All groups were tested for their abilities to remember the associated words 12 hours after learning the list. No one in the study was sleep deprived, but the subjects in two groups slept before testing and those in the other two groups did not.

The individuals who slept remembered 94% of the associations; those who did not sleep remembered 82% of the words. Those differences fell just short of statistical significance (P = .06).

The differences were marked, however, in the remaining two groups subjected to associative interference. For this part of the study, both groups also learned the same list of 20 word associations.

Again, one group slept and the other did not, and both groups were tested at 12 hours. But 12 minutes before they were tested on the list of words, both groups were subjected to associative interference. They were asked to learn a second list of 20 additional words related to the same word cues. Subjects who had slept and were introduced to the second list of words were able to recall 76% of the word associations from the first list. The subjects who had not slept were able to remember just 32% of the words. The difference in performance was highly significant (P less than .0001).

Sleep is ultimately a brain state that may hold the keys to understanding the neurobiology of memory consolidation, he said. The next step in his research will be to look at the relationships between sleep disorders, cognitive impairment, and neurologic disorders. Dr. Ellenbogen's research was supported by the National Institutes of Health and the University of Pennsylvania's Nassau Undergraduate Research Fund.