Neil Osterweil

BOSTON — Rapid thromboelastography identified trauma-induced coagulopathy within 5 minutes, produced results that correlated well with conventional coagulation tests, and predicted the early need for blood products, reported Dr. Bryan A. Cotton from a pilot study.

In the study of 272 adults treated at a trauma center over a 5-month period, the median time to view early rapid thromboelastography (rapidTEG) results, including activated clotting time, R value, and K time, was 5.2 minutes (interquartile range [IQR], 4.3-7.1 minutes), compared with 23 minutes (IQR, 18-50) for platelet count and 27 minutes (IQR, 19-69) for prothrombin time, activated partial thromboplastin time, and international normalized ratio, said Dr. Cotton of the surgery faculty in the division of acute care surgery at the University of Texas at Houston.

“Conventional labs are often quite delayed, and reflect where we were, and not where we are, in our coagulation resuscitation. RapidTEG has recently been noted as a potential solution to this. RapidTEG correlates well with conventional coagulation testing that we’re already doing, and appears to predict [need for] transfusion products in a timely fashion,” he said at the annual meeting of the American Association for the Surgery of Trauma.

In conventional thromboelastography, or TEG, coagulation is initiated with the addition of kaolin, but in rapidTEG the initiator is protein tissue factor. In 2009, Dr. Victor Jeger and colleagues from Bern (Switzerland) University Hospital published a study comparing TEG with rapidTEG in 20 patients. They found that the mean time from resuscitation bay admission until results were available for rapidTEG was 30.8 + 5.72 minutes, compared with 41.5 + 5.66 for TEG, and 64.9 + 18.8 for conventional coagulation tests (J. Trauma 2009;66:1253-7).

Based on these results, Dr. Cotton and colleagues designed a larger study to evaluate the timeliness of rapidTEG results, determine whether they correlated with conventional coagulation tests, and evaluate their ability to predict early transfusion of blood and blood components.

The prospective cohort study involved 272 patients who had the highest-level trauma activations (code 3) and were treated in the University of Texas center from October 2009 through February 2010. The cohort included scene transport patients over age 17; patients transferred from other facilities were excluded.

“Using the remote display function, early rapidTEG values were available within 5 minutes of dropping off a specimen; later values were returned within 15 minutes,” Dr. Cotton said. The late rapidTEG results, which included alpha angle, showing the rate of clot, and maximal amplitude, showing the greatest amplitude of tracing, were returned within a median of 14.9 minutes (IQR, 13.4-19.2).

The time to display rapidTEG values was significantly shorter than that for both prothrombin time and platelet counts (P less than 0.001 for both comparisons).

The investigators assessed the correlation between rapidTEG results and conventional coagulation test results, and found that rapidTEG was significantly correlated with both prothrombin time and activated partial thromboplastin time for activated clotting time, K time, R value, alpha angle, maximal amplitude, and G value. RapidTEG also correlated significantly with platelet count for K time, alpha angle, maximal amplitude, and G value.

The authors also performed a multivariate logistic regression analysis, controlling for demographic and emergency department variables, to determine whether rapidTEG could accurately predict massive transfusion (defined as more than 10 units of packed red blood cells) in the first 6 hours. They found that activated clotting time greater than 128 seconds was significantly associated with the need for transfusion, with an odds ratio of 5.15 (P = .016). Activated clotting time is the earliest rapidTEG result returned, reflecting this value’s potential utility for determining the need for blood products, Dr. Cotton said.

The investigators also found that activated clotting time could significantly predict plasma and platelet transfusion in the first 2 hours. In contrast, international normalized ratio predicted plasma transfusion but failed to predict need for either packed RBC or platelet transfusion, he said.

The invited discussant, Dr. Mitchell Cohen of San Francisco General Hospital and Trauma Center, pointed out that there were no significant differences in timing between completion of the rapidTEG tracing and return of the clinical labs, which occurred at 27 minutes after admission to the trauma bay.

“While many of us, myself included, believe that TEG will ultimately provide faster and, more importantly, more detailed information regarding which part of the coagulation system is perturbed and should be treated, there’s no data or discussion within this paper showing they used TEG in this manner,” he said.

Dr. Cohen said that answering some of the most important questions about TEG will require a randomized observational trial characterizing coagulopathy and hemostatic resuscitation after trauma, and a concurrent or separate randomized trial comparing TEG with clinical judgment and conventional coagulation testing.

There was no industry support for the study. Dr. Cotton has recently received funding for a future study of rapidTEG in trauma from Haemonetics Corp., the manufacturer of the system used in his center. Dr. Cohen said he had no conflicts of interest.

BOSTON — Rapid thromboelastography identified trauma-induced coagulopathy within 5 minutes, produced results that correlated well with conventional coagulation tests, and predicted the early need for blood products, reported Dr. Bryan A. Cotton from a pilot study.

In the study of 272 adults treated at a trauma center over a 5-month period, the median time to view early rapid thromboelastography (rapidTEG) results, including activated clotting time, R value, and K time, was 5.2 minutes (interquartile range [IQR], 4.3-7.1 minutes), compared with 23 minutes (IQR, 18-50) for platelet count and 27 minutes (IQR, 19-69) for prothrombin time, activated partial thromboplastin time, and international normalized ratio, said Dr. Cotton of the surgery faculty in the division of acute care surgery at the University of Texas at Houston.

“Conventional labs are often quite delayed, and reflect where we were, and not where we are, in our coagulation resuscitation. RapidTEG has recently been noted as a potential solution to this. RapidTEG correlates well with conventional coagulation testing that we’re already doing, and appears to predict [need for] transfusion products in a timely fashion,” he said at the annual meeting of the American Association for the Surgery of Trauma.

In conventional thromboelastography, or TEG, coagulation is initiated with the addition of kaolin, but in rapidTEG the initiator is protein tissue factor. In 2009, Dr. Victor Jeger and colleagues from Bern (Switzerland) University Hospital published a study comparing TEG with rapidTEG in 20 patients. They found that the mean time from resuscitation bay admission until results were available for rapidTEG was 30.8 + 5.72 minutes, compared with 41.5 + 5.66 for TEG, and 64.9 + 18.8 for conventional coagulation tests (J. Trauma 2009;66:1253-7).

Based on these results, Dr. Cotton and colleagues designed a larger study to evaluate the timeliness of rapidTEG results, determine whether they correlated with conventional coagulation tests, and evaluate their ability to predict early transfusion of blood and blood components.

The prospective cohort study involved 272 patients who had the highest-level trauma activations (code 3) and were treated in the University of Texas center from October 2009 through February 2010. The cohort included scene transport patients over age 17; patients transferred from other facilities were excluded.

“Using the remote display function, early rapidTEG values were available within 5 minutes of dropping off a specimen; later values were returned within 15 minutes,” Dr. Cotton said. The late rapidTEG results, which included alpha angle, showing the rate of clot, and maximal amplitude, showing the greatest amplitude of tracing, were returned within a median of 14.9 minutes (IQR, 13.4-19.2).

The time to display rapidTEG values was significantly shorter than that for both prothrombin time and platelet counts (P less than 0.001 for both comparisons).

The investigators assessed the correlation between rapidTEG results and conventional coagulation test results, and found that rapidTEG was significantly correlated with both prothrombin time and activated partial thromboplastin time for activated clotting time, K time, R value, alpha angle, maximal amplitude, and G value. RapidTEG also correlated significantly with platelet count for K time, alpha angle, maximal amplitude, and G value.

The authors also performed a multivariate logistic regression analysis, controlling for demographic and emergency department variables, to determine whether rapidTEG could accurately predict massive transfusion (defined as more than 10 units of packed red blood cells) in the first 6 hours. They found that activated clotting time greater than 128 seconds was significantly associated with the need for transfusion, with an odds ratio of 5.15 (P = .016). Activated clotting time is the earliest rapidTEG result returned, reflecting this value’s potential utility for determining the need for blood products, Dr. Cotton said.

The investigators also found that activated clotting time could significantly predict plasma and platelet transfusion in the first 2 hours. In contrast, international normalized ratio predicted plasma transfusion but failed to predict need for either packed RBC or platelet transfusion, he said.

The invited discussant, Dr. Mitchell Cohen of San Francisco General Hospital and Trauma Center, pointed out that there were no significant differences in timing between completion of the rapidTEG tracing and return of the clinical labs, which occurred at 27 minutes after admission to the trauma bay.

“While many of us, myself included, believe that TEG will ultimately provide faster and, more importantly, more detailed information regarding which part of the coagulation system is perturbed and should be treated, there’s no data or discussion within this paper showing they used TEG in this manner,” he said.

Dr. Cohen said that answering some of the most important questions about TEG will require a randomized observational trial characterizing coagulopathy and hemostatic resuscitation after trauma, and a concurrent or separate randomized trial comparing TEG with clinical judgment and conventional coagulation testing.

There was no industry support for the study. Dr. Cotton has recently received funding for a future study of rapidTEG in trauma from Haemonetics Corp., the manufacturer of the system used in his center. Dr. Cohen said he had no conflicts of interest.

BOSTON — Rapid thromboelastography identified trauma-induced coagulopathy within 5 minutes, produced results that correlated well with conventional coagulation tests, and predicted the early need for blood products, reported Dr. Bryan A. Cotton from a pilot study.

In the study of 272 adults treated at a trauma center over a 5-month period, the median time to view early rapid thromboelastography (rapidTEG) results, including activated clotting time, R value, and K time, was 5.2 minutes (interquartile range [IQR], 4.3-7.1 minutes), compared with 23 minutes (IQR, 18-50) for platelet count and 27 minutes (IQR, 19-69) for prothrombin time, activated partial thromboplastin time, and international normalized ratio, said Dr. Cotton of the surgery faculty in the division of acute care surgery at the University of Texas at Houston.

“Conventional labs are often quite delayed, and reflect where we were, and not where we are, in our coagulation resuscitation. RapidTEG has recently been noted as a potential solution to this. RapidTEG correlates well with conventional coagulation testing that we’re already doing, and appears to predict [need for] transfusion products in a timely fashion,” he said at the annual meeting of the American Association for the Surgery of Trauma.

In conventional thromboelastography, or TEG, coagulation is initiated with the addition of kaolin, but in rapidTEG the initiator is protein tissue factor. In 2009, Dr. Victor Jeger and colleagues from Bern (Switzerland) University Hospital published a study comparing TEG with rapidTEG in 20 patients. They found that the mean time from resuscitation bay admission until results were available for rapidTEG was 30.8 + 5.72 minutes, compared with 41.5 + 5.66 for TEG, and 64.9 + 18.8 for conventional coagulation tests (J. Trauma 2009;66:1253-7).

Based on these results, Dr. Cotton and colleagues designed a larger study to evaluate the timeliness of rapidTEG results, determine whether they correlated with conventional coagulation tests, and evaluate their ability to predict early transfusion of blood and blood components.

The prospective cohort study involved 272 patients who had the highest-level trauma activations (code 3) and were treated in the University of Texas center from October 2009 through February 2010. The cohort included scene transport patients over age 17; patients transferred from other facilities were excluded.

“Using the remote display function, early rapidTEG values were available within 5 minutes of dropping off a specimen; later values were returned within 15 minutes,” Dr. Cotton said. The late rapidTEG results, which included alpha angle, showing the rate of clot, and maximal amplitude, showing the greatest amplitude of tracing, were returned within a median of 14.9 minutes (IQR, 13.4-19.2).

The time to display rapidTEG values was significantly shorter than that for both prothrombin time and platelet counts (P less than 0.001 for both comparisons).

The investigators assessed the correlation between rapidTEG results and conventional coagulation test results, and found that rapidTEG was significantly correlated with both prothrombin time and activated partial thromboplastin time for activated clotting time, K time, R value, alpha angle, maximal amplitude, and G value. RapidTEG also correlated significantly with platelet count for K time, alpha angle, maximal amplitude, and G value.

The authors also performed a multivariate logistic regression analysis, controlling for demographic and emergency department variables, to determine whether rapidTEG could accurately predict massive transfusion (defined as more than 10 units of packed red blood cells) in the first 6 hours. They found that activated clotting time greater than 128 seconds was significantly associated with the need for transfusion, with an odds ratio of 5.15 (P = .016). Activated clotting time is the earliest rapidTEG result returned, reflecting this value’s potential utility for determining the need for blood products, Dr. Cotton said.

The investigators also found that activated clotting time could significantly predict plasma and platelet transfusion in the first 2 hours. In contrast, international normalized ratio predicted plasma transfusion but failed to predict need for either packed RBC or platelet transfusion, he said.

The invited discussant, Dr. Mitchell Cohen of San Francisco General Hospital and Trauma Center, pointed out that there were no significant differences in timing between completion of the rapidTEG tracing and return of the clinical labs, which occurred at 27 minutes after admission to the trauma bay.

“While many of us, myself included, believe that TEG will ultimately provide faster and, more importantly, more detailed information regarding which part of the coagulation system is perturbed and should be treated, there’s no data or discussion within this paper showing they used TEG in this manner,” he said.

Dr. Cohen said that answering some of the most important questions about TEG will require a randomized observational trial characterizing coagulopathy and hemostatic resuscitation after trauma, and a concurrent or separate randomized trial comparing TEG with clinical judgment and conventional coagulation testing.

There was no industry support for the study. Dr. Cotton has recently received funding for a future study of rapidTEG in trauma from Haemonetics Corp., the manufacturer of the system used in his center. Dr. Cohen said he had no conflicts of interest.

BOSTON – In critically injured patients, early treatment with vasopressors was associated with more than an 11-fold increase in risk of death, and this risk was independent of the patient’s fluid status, reported Dr. David Plurad.

An analysis of outcomes of 1,349 patients treated over an 8-year period showed that among all patients who received vasopressors within 24 hours of admission, those with low scores (less than 8) on the Glasgow Coma Scale (GCS) had a more than fourfold increased risk for death, and those who received more than one vasopressor had a risk nearly as large, reported Dr. Plurad of the Los Angeles County Hospital/University of Southern California Medical Center.

“While we know that these vasopressors are necessary many times, they should be used with caution regardless of the presumed adequacy of resuscitation. Further investigation is necessary to develop evidence-based strategies that maximize beneficial, and minimize detrimental, effects when these agents become necessary after trauma,” he said at the annual meeting of the American Association for the Surgery of Trauma.

In trauma cases, vasopressors are typically used early in the course of treatment when a patient experiences traumatic arrest requiring thoracotomy or becomes hypotensive during surgery while the team is attempting to control bleeding, or during damage-control resuscitation while the team attempts to correct physiologic parameters.

The use of vasopressors is decades old, as are warnings to clinicians about their indiscriminate use. But recent evidence from animal studies, short human clinical trials, and case reports suggests that vasopressors might be useful for maintaining critical central perfusion and delaying cardiac arrest because of hemorrhaging, Dr. Plurad said.

To study whether vasopressors were associated with increased risk of death, the investigators conducted a retrospective database/registry search on all trauma patients older than 15 years admitted to their ICU in 2001-2008 who had central venous pressure monitoring. They excluded from their analyses all patients with traumatic brain injury (defined as an Abbreviated Injury Scale score greater than 3), and all patients with spinal cord injuries.

The two-stage analysis looked at all patients, and the subset of patients who received vasopressors during the first 24 hours after ICU admission. The agents included dopamine, epinephrine, norepinephrine, vasopressin, and phenylephrine. Patients who received dobutamine or milrinone were included in the vasopressor-negative group.

Of the 1,349 total patients – nearly 80% of whom were male victims of blunt trauma – 351 (26.0%) received at least one vasopressor, and 150 (11.1%) received multiple agents. The most commonly used vasopressor was dopamine, in 22.5% of all patients, followed by norepinephrine in 6.4%, epinephrine in 5.9%, vasopressin in 4.4%, and phenylephrine in 2.4%. Overall, 195 patients (14.5%) died.

Patients who received vasopressin were significantly more likely than those who did not to have blunt mechanism of trauma (odds ratio [OR], 1.98), to be admitted to the ICU with hypovolemia (OR, 9.55), to have an emergency department GCS score of 8 or less (OR, 3.7), and to die in the hospital (OR, 17.6); they were also older and had higher mean injury severity scores.

Factors independently associated with death among all patients were vasopressor use (OR, 11.51), emergency department GCS score of 8 or less (OR, 4.1), injury severity score of 35 or greater (OR, 2.71), and age of 55 or older (OR, 2.3; P value for all comparisons less than .01).

In logistic regression analysis of all patients who received vasopressors, an emergency department GCS score of 8 or lower was significantly associated with mortality (OR, 4.33, P less than .01), as was multiple vasopressor use (OR, 3.93, P less than 0.01). In contrast, hypovolemia was not significantly associated with increased risk of death in these patients (OR, 1.29).

Commenting on the study, invited discussant Dr. Patrick Offner of St. Anthony Central Hospital in Denver noted that trauma guidelines emphasize that vasopressors should be used with caution in trauma patients.

“This is, however, a retrospective study, and as such, suffers from several limitations,” he said. “In particular, the specific indications and timing of vasopressor use are unknown.”

He added that although hypovolemic and nonhypovolemic patients in the study seemed to be relatively well matched, with similar anatomic injury severity scores and rates of hypotension, potentially important physiologic information was missing.

“Ultimately, I think it remains unclear whether early vasopressor use is in itself detrimental or merely a marker for outcome in these patients,” Dr. Offner said.

The funding source for the study was not disclosed. Dr. Offner reported that he had no conflicts of interest.

BOSTON – In critically injured patients, early treatment with vasopressors was associated with more than an 11-fold increase in risk of death, and this risk was independent of the patient’s fluid status, reported Dr. David Plurad.

An analysis of outcomes of 1,349 patients treated over an 8-year period showed that among all patients who received vasopressors within 24 hours of admission, those with low scores (less than 8) on the Glasgow Coma Scale (GCS) had a more than fourfold increased risk for death, and those who received more than one vasopressor had a risk nearly as large, reported Dr. Plurad of the Los Angeles County Hospital/University of Southern California Medical Center.

“While we know that these vasopressors are necessary many times, they should be used with caution regardless of the presumed adequacy of resuscitation. Further investigation is necessary to develop evidence-based strategies that maximize beneficial, and minimize detrimental, effects when these agents become necessary after trauma,” he said at the annual meeting of the American Association for the Surgery of Trauma.

In trauma cases, vasopressors are typically used early in the course of treatment when a patient experiences traumatic arrest requiring thoracotomy or becomes hypotensive during surgery while the team is attempting to control bleeding, or during damage-control resuscitation while the team attempts to correct physiologic parameters.

The use of vasopressors is decades old, as are warnings to clinicians about their indiscriminate use. But recent evidence from animal studies, short human clinical trials, and case reports suggests that vasopressors might be useful for maintaining critical central perfusion and delaying cardiac arrest because of hemorrhaging, Dr. Plurad said.

To study whether vasopressors were associated with increased risk of death, the investigators conducted a retrospective database/registry search on all trauma patients older than 15 years admitted to their ICU in 2001-2008 who had central venous pressure monitoring. They excluded from their analyses all patients with traumatic brain injury (defined as an Abbreviated Injury Scale score greater than 3), and all patients with spinal cord injuries.

The two-stage analysis looked at all patients, and the subset of patients who received vasopressors during the first 24 hours after ICU admission. The agents included dopamine, epinephrine, norepinephrine, vasopressin, and phenylephrine. Patients who received dobutamine or milrinone were included in the vasopressor-negative group.

Of the 1,349 total patients – nearly 80% of whom were male victims of blunt trauma – 351 (26.0%) received at least one vasopressor, and 150 (11.1%) received multiple agents. The most commonly used vasopressor was dopamine, in 22.5% of all patients, followed by norepinephrine in 6.4%, epinephrine in 5.9%, vasopressin in 4.4%, and phenylephrine in 2.4%. Overall, 195 patients (14.5%) died.

Patients who received vasopressin were significantly more likely than those who did not to have blunt mechanism of trauma (odds ratio [OR], 1.98), to be admitted to the ICU with hypovolemia (OR, 9.55), to have an emergency department GCS score of 8 or less (OR, 3.7), and to die in the hospital (OR, 17.6); they were also older and had higher mean injury severity scores.

Factors independently associated with death among all patients were vasopressor use (OR, 11.51), emergency department GCS score of 8 or less (OR, 4.1), injury severity score of 35 or greater (OR, 2.71), and age of 55 or older (OR, 2.3; P value for all comparisons less than .01).

In logistic regression analysis of all patients who received vasopressors, an emergency department GCS score of 8 or lower was significantly associated with mortality (OR, 4.33, P less than .01), as was multiple vasopressor use (OR, 3.93, P less than 0.01). In contrast, hypovolemia was not significantly associated with increased risk of death in these patients (OR, 1.29).

Commenting on the study, invited discussant Dr. Patrick Offner of St. Anthony Central Hospital in Denver noted that trauma guidelines emphasize that vasopressors should be used with caution in trauma patients.

“This is, however, a retrospective study, and as such, suffers from several limitations,” he said. “In particular, the specific indications and timing of vasopressor use are unknown.”

He added that although hypovolemic and nonhypovolemic patients in the study seemed to be relatively well matched, with similar anatomic injury severity scores and rates of hypotension, potentially important physiologic information was missing.

“Ultimately, I think it remains unclear whether early vasopressor use is in itself detrimental or merely a marker for outcome in these patients,” Dr. Offner said.

The funding source for the study was not disclosed. Dr. Offner reported that he had no conflicts of interest.

BOSTON – In critically injured patients, early treatment with vasopressors was associated with more than an 11-fold increase in risk of death, and this risk was independent of the patient’s fluid status, reported Dr. David Plurad.

An analysis of outcomes of 1,349 patients treated over an 8-year period showed that among all patients who received vasopressors within 24 hours of admission, those with low scores (less than 8) on the Glasgow Coma Scale (GCS) had a more than fourfold increased risk for death, and those who received more than one vasopressor had a risk nearly as large, reported Dr. Plurad of the Los Angeles County Hospital/University of Southern California Medical Center.

“While we know that these vasopressors are necessary many times, they should be used with caution regardless of the presumed adequacy of resuscitation. Further investigation is necessary to develop evidence-based strategies that maximize beneficial, and minimize detrimental, effects when these agents become necessary after trauma,” he said at the annual meeting of the American Association for the Surgery of Trauma.

In trauma cases, vasopressors are typically used early in the course of treatment when a patient experiences traumatic arrest requiring thoracotomy or becomes hypotensive during surgery while the team is attempting to control bleeding, or during damage-control resuscitation while the team attempts to correct physiologic parameters.

The use of vasopressors is decades old, as are warnings to clinicians about their indiscriminate use. But recent evidence from animal studies, short human clinical trials, and case reports suggests that vasopressors might be useful for maintaining critical central perfusion and delaying cardiac arrest because of hemorrhaging, Dr. Plurad said.

To study whether vasopressors were associated with increased risk of death, the investigators conducted a retrospective database/registry search on all trauma patients older than 15 years admitted to their ICU in 2001-2008 who had central venous pressure monitoring. They excluded from their analyses all patients with traumatic brain injury (defined as an Abbreviated Injury Scale score greater than 3), and all patients with spinal cord injuries.

The two-stage analysis looked at all patients, and the subset of patients who received vasopressors during the first 24 hours after ICU admission. The agents included dopamine, epinephrine, norepinephrine, vasopressin, and phenylephrine. Patients who received dobutamine or milrinone were included in the vasopressor-negative group.

Of the 1,349 total patients – nearly 80% of whom were male victims of blunt trauma – 351 (26.0%) received at least one vasopressor, and 150 (11.1%) received multiple agents. The most commonly used vasopressor was dopamine, in 22.5% of all patients, followed by norepinephrine in 6.4%, epinephrine in 5.9%, vasopressin in 4.4%, and phenylephrine in 2.4%. Overall, 195 patients (14.5%) died.

Patients who received vasopressin were significantly more likely than those who did not to have blunt mechanism of trauma (odds ratio [OR], 1.98), to be admitted to the ICU with hypovolemia (OR, 9.55), to have an emergency department GCS score of 8 or less (OR, 3.7), and to die in the hospital (OR, 17.6); they were also older and had higher mean injury severity scores.

Factors independently associated with death among all patients were vasopressor use (OR, 11.51), emergency department GCS score of 8 or less (OR, 4.1), injury severity score of 35 or greater (OR, 2.71), and age of 55 or older (OR, 2.3; P value for all comparisons less than .01).

In logistic regression analysis of all patients who received vasopressors, an emergency department GCS score of 8 or lower was significantly associated with mortality (OR, 4.33, P less than .01), as was multiple vasopressor use (OR, 3.93, P less than 0.01). In contrast, hypovolemia was not significantly associated with increased risk of death in these patients (OR, 1.29).

Commenting on the study, invited discussant Dr. Patrick Offner of St. Anthony Central Hospital in Denver noted that trauma guidelines emphasize that vasopressors should be used with caution in trauma patients.

“This is, however, a retrospective study, and as such, suffers from several limitations,” he said. “In particular, the specific indications and timing of vasopressor use are unknown.”

He added that although hypovolemic and nonhypovolemic patients in the study seemed to be relatively well matched, with similar anatomic injury severity scores and rates of hypotension, potentially important physiologic information was missing.

“Ultimately, I think it remains unclear whether early vasopressor use is in itself detrimental or merely a marker for outcome in these patients,” Dr. Offner said.

The funding source for the study was not disclosed. Dr. Offner reported that he had no conflicts of interest.

BOSTON  – A third of patients transferred to a level 1 regional trauma center were sent there unnecessarily and at substantial cost to the center, and most of this “secondary overtriage” could be attributed to nonmedical reasons such as time of day or the patient’s insurance status rather than severity of injury.

Of 2,486 patients transferred, 374 who were admitted to the trauma center had injuries that could have been safely handled by the referring institutions, and 582 were seen and released from the emergency department, Dr. Eric A. Toschlog said at the American Association for the Surgery of Trauma Annual Meeting.

From 2007 through 2009, secondary overtriage resulted in an estimated $570,000 loss for the center, whereas appropriate transfers, after adjustment for Medicaid payments, resulted in an estimated $1.5 million gain.

“Overtriage is a costly entity to our trauma center,” said Dr. Toschlog, of the Brody School of Medicine at East Carolina University in Greenville, N.C.

The study results, he said, are in agreement with the American College of Surgeons’ Resources for Optimal Care of the Injured Patient (Green Book), which states that overtriage has no or at most minimal consequences for patients, but results in “excessive costs and burden for higher level trauma centers in the routine care of injured patients.”

Despite that assertion, overtriage rates as high as 50% have been deemed acceptable as a means of preventing undertriage, Dr. Toschlog said.

The term “secondary overtriage” was coined by investigators from the Washington Hospital Center in Washington as overtriage originating from outlying hospitals (J. Am. Coll. Surg. 2008;206:131-7).

To determine the secondary overtriage rate, identify risk factors associated with transfer, and evaluate its financial impact, Dr. Toschlog and colleagues retrospectively reviewed data from their trauma registry on consecutive trauma transfer admissions over 24 months from 2007 to 2009.

They used Green Book criteria to define appropriate triage as death at the trauma center, admission to an intensive care unit, urgent surgery, or trauma center length of stay greater than 48 hours.

The secondary overtriage rate was calculated as the difference between the total number of transfer admissions and appropriately triaged cases, divided by the total number of transfer admissions.

The appropriate and overtriage groups were compared by demographics (including insurance status), injury severity, time of transfer, and day of week. Financial variables included total, direct, and indirect hospital costs, as well as total payments before and after Medicaid year-end lump-sum adjustments to the hospital for the percentage of Medicaid and Medicare patients treated.

Of the 3,661 total admissions, 2,486, or just under 68%, were transfers from 1 of 36 facilities; 88% of the transfers came from 1 of 13 hospitals.

A total of 374 patients met the secondary overtriage definition, and 2,112 were determined to be secondary appropriate transfers. An additional 582 patients transferred from outlying hospitals were seen in the emergency department and released. The total secondary overtriage rate, therefore, was 31.2% (956 of 3,068) patients, Dr. Toschlog said.

The transfer population had a mean age of 40.7 years, 71% were male, and 56% were white. About half of the patients (52%) arrived on a weekend. The mean Injury Severity Score (ISS) was 13.9 (moderate).

Univariate analysis indicated that the overtriaged patients were more likely to be younger, to be uninsured, to arrive at night, and to be less severely injured across all measures of severity, compared with appropriately triaged patients. The overtriaged group had a mean hospital length of stay of 1 day, compared with 8.4 days for the appropriately triaged patients. Gender, ethnicity, weekend transfer, and substance abuse were not significantly associated with overtriage.

In a multivariate analysis, factors that were significantly associated with overtriage were age (odds ratio, 0.71), self-pay (OR, 1.88), night-time transfer (OR, 1.32), nonblunt trauma (OR, 0.56), ISS, and Revised Trauma Score less than 7.

As expected based on length of stay, total per-patient charges and costs for the appropriately triaged/transferred patients were higher.

Before adjustment for state Medicaid reimbursements, all transferred patients resulted in a financial loss for the trauma center. The loss for overtriaged patients averaged $1,520, and for appropriate transfers was $2,196. But after Medicaid payments were added, overtriage resulted in a $1,356 loss, whereas appropriate triage resulted in a $692 gain.

“Of particular concern is that the appropriately transferred cohort would have lost $4.6 million without government support,” Dr. Toschlog commented.

The study was limited by the retrospective design and by the possibility that the data may be unique to their institution and to their regional system, which defaults to the referring provider for acceptance of trauma transfer without input from the trauma surgeon. Overtriage criteria may have been unfairly applied, because some of the referring hospitals are not well equipped for evaluating trauma patients, he acknowledged.

“Our hope is that our study will contribute to the movement of the national trauma agenda toward further and better trauma system development. We can certainly maintain our commitment to caring for the injured while also working to ensure that the right patient is cared for in the right place at the right time,” he said.

Dr. C. William Schwab of the University of Pennsylvania, Philadelphia, the invited discussant, said that the focus of the study was too narrow, because it looked only at patient characteristics as a predictor of transfer and did not consider local hospital resources, provider level and skill complement, or distance from the local hospital to the level 1 center.

The study’s funding source was not disclosed. Dr. Toschlog and Dr. Schwab disclosed no conflicts of interest.

BOSTON  – A third of patients transferred to a level 1 regional trauma center were sent there unnecessarily and at substantial cost to the center, and most of this “secondary overtriage” could be attributed to nonmedical reasons such as time of day or the patient’s insurance status rather than severity of injury.

Of 2,486 patients transferred, 374 who were admitted to the trauma center had injuries that could have been safely handled by the referring institutions, and 582 were seen and released from the emergency department, Dr. Eric A. Toschlog said at the American Association for the Surgery of Trauma Annual Meeting.

From 2007 through 2009, secondary overtriage resulted in an estimated $570,000 loss for the center, whereas appropriate transfers, after adjustment for Medicaid payments, resulted in an estimated $1.5 million gain.

“Overtriage is a costly entity to our trauma center,” said Dr. Toschlog, of the Brody School of Medicine at East Carolina University in Greenville, N.C.

The study results, he said, are in agreement with the American College of Surgeons’ Resources for Optimal Care of the Injured Patient (Green Book), which states that overtriage has no or at most minimal consequences for patients, but results in “excessive costs and burden for higher level trauma centers in the routine care of injured patients.”

Despite that assertion, overtriage rates as high as 50% have been deemed acceptable as a means of preventing undertriage, Dr. Toschlog said.

The term “secondary overtriage” was coined by investigators from the Washington Hospital Center in Washington as overtriage originating from outlying hospitals (J. Am. Coll. Surg. 2008;206:131-7).

To determine the secondary overtriage rate, identify risk factors associated with transfer, and evaluate its financial impact, Dr. Toschlog and colleagues retrospectively reviewed data from their trauma registry on consecutive trauma transfer admissions over 24 months from 2007 to 2009.

They used Green Book criteria to define appropriate triage as death at the trauma center, admission to an intensive care unit, urgent surgery, or trauma center length of stay greater than 48 hours.

The secondary overtriage rate was calculated as the difference between the total number of transfer admissions and appropriately triaged cases, divided by the total number of transfer admissions.

The appropriate and overtriage groups were compared by demographics (including insurance status), injury severity, time of transfer, and day of week. Financial variables included total, direct, and indirect hospital costs, as well as total payments before and after Medicaid year-end lump-sum adjustments to the hospital for the percentage of Medicaid and Medicare patients treated.

Of the 3,661 total admissions, 2,486, or just under 68%, were transfers from 1 of 36 facilities; 88% of the transfers came from 1 of 13 hospitals.

A total of 374 patients met the secondary overtriage definition, and 2,112 were determined to be secondary appropriate transfers. An additional 582 patients transferred from outlying hospitals were seen in the emergency department and released. The total secondary overtriage rate, therefore, was 31.2% (956 of 3,068) patients, Dr. Toschlog said.

The transfer population had a mean age of 40.7 years, 71% were male, and 56% were white. About half of the patients (52%) arrived on a weekend. The mean Injury Severity Score (ISS) was 13.9 (moderate).

Univariate analysis indicated that the overtriaged patients were more likely to be younger, to be uninsured, to arrive at night, and to be less severely injured across all measures of severity, compared with appropriately triaged patients. The overtriaged group had a mean hospital length of stay of 1 day, compared with 8.4 days for the appropriately triaged patients. Gender, ethnicity, weekend transfer, and substance abuse were not significantly associated with overtriage.

In a multivariate analysis, factors that were significantly associated with overtriage were age (odds ratio, 0.71), self-pay (OR, 1.88), night-time transfer (OR, 1.32), nonblunt trauma (OR, 0.56), ISS, and Revised Trauma Score less than 7.

As expected based on length of stay, total per-patient charges and costs for the appropriately triaged/transferred patients were higher.

Before adjustment for state Medicaid reimbursements, all transferred patients resulted in a financial loss for the trauma center. The loss for overtriaged patients averaged $1,520, and for appropriate transfers was $2,196. But after Medicaid payments were added, overtriage resulted in a $1,356 loss, whereas appropriate triage resulted in a $692 gain.

“Of particular concern is that the appropriately transferred cohort would have lost $4.6 million without government support,” Dr. Toschlog commented.

The study was limited by the retrospective design and by the possibility that the data may be unique to their institution and to their regional system, which defaults to the referring provider for acceptance of trauma transfer without input from the trauma surgeon. Overtriage criteria may have been unfairly applied, because some of the referring hospitals are not well equipped for evaluating trauma patients, he acknowledged.

“Our hope is that our study will contribute to the movement of the national trauma agenda toward further and better trauma system development. We can certainly maintain our commitment to caring for the injured while also working to ensure that the right patient is cared for in the right place at the right time,” he said.

Dr. C. William Schwab of the University of Pennsylvania, Philadelphia, the invited discussant, said that the focus of the study was too narrow, because it looked only at patient characteristics as a predictor of transfer and did not consider local hospital resources, provider level and skill complement, or distance from the local hospital to the level 1 center.

The study’s funding source was not disclosed. Dr. Toschlog and Dr. Schwab disclosed no conflicts of interest.

BOSTON  – A third of patients transferred to a level 1 regional trauma center were sent there unnecessarily and at substantial cost to the center, and most of this “secondary overtriage” could be attributed to nonmedical reasons such as time of day or the patient’s insurance status rather than severity of injury.

Of 2,486 patients transferred, 374 who were admitted to the trauma center had injuries that could have been safely handled by the referring institutions, and 582 were seen and released from the emergency department, Dr. Eric A. Toschlog said at the American Association for the Surgery of Trauma Annual Meeting.

From 2007 through 2009, secondary overtriage resulted in an estimated $570,000 loss for the center, whereas appropriate transfers, after adjustment for Medicaid payments, resulted in an estimated $1.5 million gain.

“Overtriage is a costly entity to our trauma center,” said Dr. Toschlog, of the Brody School of Medicine at East Carolina University in Greenville, N.C.

The study results, he said, are in agreement with the American College of Surgeons’ Resources for Optimal Care of the Injured Patient (Green Book), which states that overtriage has no or at most minimal consequences for patients, but results in “excessive costs and burden for higher level trauma centers in the routine care of injured patients.”

Despite that assertion, overtriage rates as high as 50% have been deemed acceptable as a means of preventing undertriage, Dr. Toschlog said.

The term “secondary overtriage” was coined by investigators from the Washington Hospital Center in Washington as overtriage originating from outlying hospitals (J. Am. Coll. Surg. 2008;206:131-7).

To determine the secondary overtriage rate, identify risk factors associated with transfer, and evaluate its financial impact, Dr. Toschlog and colleagues retrospectively reviewed data from their trauma registry on consecutive trauma transfer admissions over 24 months from 2007 to 2009.

They used Green Book criteria to define appropriate triage as death at the trauma center, admission to an intensive care unit, urgent surgery, or trauma center length of stay greater than 48 hours.

The secondary overtriage rate was calculated as the difference between the total number of transfer admissions and appropriately triaged cases, divided by the total number of transfer admissions.

The appropriate and overtriage groups were compared by demographics (including insurance status), injury severity, time of transfer, and day of week. Financial variables included total, direct, and indirect hospital costs, as well as total payments before and after Medicaid year-end lump-sum adjustments to the hospital for the percentage of Medicaid and Medicare patients treated.

Of the 3,661 total admissions, 2,486, or just under 68%, were transfers from 1 of 36 facilities; 88% of the transfers came from 1 of 13 hospitals.

A total of 374 patients met the secondary overtriage definition, and 2,112 were determined to be secondary appropriate transfers. An additional 582 patients transferred from outlying hospitals were seen in the emergency department and released. The total secondary overtriage rate, therefore, was 31.2% (956 of 3,068) patients, Dr. Toschlog said.

The transfer population had a mean age of 40.7 years, 71% were male, and 56% were white. About half of the patients (52%) arrived on a weekend. The mean Injury Severity Score (ISS) was 13.9 (moderate).

Univariate analysis indicated that the overtriaged patients were more likely to be younger, to be uninsured, to arrive at night, and to be less severely injured across all measures of severity, compared with appropriately triaged patients. The overtriaged group had a mean hospital length of stay of 1 day, compared with 8.4 days for the appropriately triaged patients. Gender, ethnicity, weekend transfer, and substance abuse were not significantly associated with overtriage.

In a multivariate analysis, factors that were significantly associated with overtriage were age (odds ratio, 0.71), self-pay (OR, 1.88), night-time transfer (OR, 1.32), nonblunt trauma (OR, 0.56), ISS, and Revised Trauma Score less than 7.

As expected based on length of stay, total per-patient charges and costs for the appropriately triaged/transferred patients were higher.

Before adjustment for state Medicaid reimbursements, all transferred patients resulted in a financial loss for the trauma center. The loss for overtriaged patients averaged $1,520, and for appropriate transfers was $2,196. But after Medicaid payments were added, overtriage resulted in a $1,356 loss, whereas appropriate triage resulted in a $692 gain.

“Of particular concern is that the appropriately transferred cohort would have lost $4.6 million without government support,” Dr. Toschlog commented.

The study was limited by the retrospective design and by the possibility that the data may be unique to their institution and to their regional system, which defaults to the referring provider for acceptance of trauma transfer without input from the trauma surgeon. Overtriage criteria may have been unfairly applied, because some of the referring hospitals are not well equipped for evaluating trauma patients, he acknowledged.

“Our hope is that our study will contribute to the movement of the national trauma agenda toward further and better trauma system development. We can certainly maintain our commitment to caring for the injured while also working to ensure that the right patient is cared for in the right place at the right time,” he said.

Dr. C. William Schwab of the University of Pennsylvania, Philadelphia, the invited discussant, said that the focus of the study was too narrow, because it looked only at patient characteristics as a predictor of transfer and did not consider local hospital resources, provider level and skill complement, or distance from the local hospital to the level 1 center.

The study’s funding source was not disclosed. Dr. Toschlog and Dr. Schwab disclosed no conflicts of interest.

BOSTON – Venous thromboembolism is a common complication of traumatic brain injury, and pharmacologic thromboprophylaxis is considered to be the standard of care. But when that prophylaxis is interrupted for surgery or other procedures, the patient’s risk for VTE significantly increases, reported investigators at the annual meeting of the American Association for the Surgery of Trauma.

Among 480 patients with stable traumatic brain injury (TBI), those who received interrupted pharmacologic thromboprophylaxis (PTP) had a sevenfold higher risk for VTE than did patients who received continuous prophylaxis, said Dr. Patrick Offner of the trauma service at St. Anthony Central Hospital in Denver.

“We would suggest that interruption of PTP in patients with traumatic brain injury does increase the risk of venous thromboembolism, and all efforts should be made to not interrupt it once it’s started,” Dr. Offner said.

Patients with traumatic brain injury are at increased risk for VTE because of multiple mechanisms, some related to injury and some to treatment. But efforts to prevent VTE in these patients are particularly challenging, because prophylaxis with an agent such as low-molecular-weight heparin also increases risk of intracranial hemorrhage, he said.

Dr. Offner and colleagues recently published a study suggesting that PTP use is associated with a 13-fold increase in odds of further hemorrhage progression in patients with unstable TBI, defined as evidence of intracranial hemorrhage progression on a follow-up CT scan within 1 day of admission (J. Trauma 2010;68:886-94).

To get a better handle on the optimal administration of PTP in patients with stable TBI and to see whether late or interrupted PTP might increase the risk of VTE, the investigators conducted a retrospective cohort study.

They reviewed records on all adult patients admitted to two urban level 1 trauma centers with a diagnosis of TBI based on ICD-9 codes. Patients with hospital stays shorter than 3 days, new or aggressive intracranial hemorrhage within 24 hours, or VTE diagnosed within 24 hours of admission were excluded.

The VTE prophylaxis protocol included placement of bilateral sequential compression devices (unless precluded by lower-extremity injuries) on admission to the ICU, and Lovenox (enoxaparin) 30 mg subcutaneously every 12 hours beginning at 36 hours, unless it was specifically held by neurosurgeons or other members of the care team. The patients also received weekly duplex ultrasound surveillance of the lower extremities.

The authors collected data on development of deep vein thrombosis (DVT) or pulmonary embolism (PE), DVT prophylaxis timing and method, and demographics. The analysis included incidence of VTE, associations between VTE and PTP – any exposure, early exposure (within 72 hours) compared with late, and interrupted administration of PTP – and multivariate logistic regression modeling.

Information on a total of 480 patients was available. The median age was 53 years, 62% were men, and 48% of injuries were due to falls. The mean Injury Severity Score was 19.7, mean Abbreviated Injury Scale (AIS) score was 3.5, and mean Glasgow Coma Scale (GCS) score was 12.2.

There were 15 venous thromboembolic events: 12 DVT and 5 PE (two patients had DVT and PE). The incidence was lower than the investigators had anticipated, Dr. Offner said.

In all, 93% of the patients were treated with sequential compression, and 255 received PTP. Of these, 42% received PTP early, 58% received it late, 74% received it continuously, and 26% had PTP interrupted.

Factors significantly related to the subsequent development of VTE included a GCS score of 3-8, compared with 9-15 for patients who did not develop a VTE; severe head injury (AIS score = 5); chest or extremity injury (AIS score 3 or greater for each); and less ambulation before discharge.

“When we looked at the univariate association between VTE and PTP, we were surprised in that administration of PTP either at all or in a late or early fashion didn’t seem to affect the subsequent incidence of VTE. On the other hand, if PTP was interrupted after it was started, there was a fourfold increase [odds ratio, 4.53; P = .02] in the incidence of subsequent venous thromboembolism,” Dr. Offner said.

In logistic regression analyses controlling for various risk factors, the investigators first compared PTP with no PTP and VTE incidence, and found that the only significant risk factor was ambulation before discharge (OR, 0.19; P = .02). Severity of injuries, preinjury anticoagulant/platelet inhibitor use, low GCS score, or female sex were not significantly linked to VTE incidence, regardless of PTP use.

There were also no significant differences in VTE incidence in late, compared with early administration of PTP.

However, when they looked at interrupted compared with continuous administration, they found that interrupted administration was associated with a sevenfold increased risk for VTE (OR, 7.07; P = .04). None of the other variables were significantly linked to VTE risk in this subanalysis.

Dr. Offner noted that the study results were limited by the retrospective design, potential for surveillance bias, and relatively small number of VTE cases in the sample.

Invited discussant Dr. Thomas Esposito of the department of surgery at Loyola University Medical Center in Maywood, Ill., noted that there was a mismatch between the interrupted and continuous PTP administration groups. That called the data into question, he said, because differences between the groups might have significantly influenced the incidence of VTE and PE.

Dr. Offner acknowledged that the small numbers involved could have subjected the study to a type II statistical error. He also agreed that the interrupted therapy group was a higher-risk group and that those risk factors could in part explain the higher incidence of VTE.

The study was internally funded. Neither Dr. Offner nor Dr. Esposito had conflict of interest disclosures.

BOSTON – Venous thromboembolism is a common complication of traumatic brain injury, and pharmacologic thromboprophylaxis is considered to be the standard of care. But when that prophylaxis is interrupted for surgery or other procedures, the patient’s risk for VTE significantly increases, reported investigators at the annual meeting of the American Association for the Surgery of Trauma.

Among 480 patients with stable traumatic brain injury (TBI), those who received interrupted pharmacologic thromboprophylaxis (PTP) had a sevenfold higher risk for VTE than did patients who received continuous prophylaxis, said Dr. Patrick Offner of the trauma service at St. Anthony Central Hospital in Denver.

“We would suggest that interruption of PTP in patients with traumatic brain injury does increase the risk of venous thromboembolism, and all efforts should be made to not interrupt it once it’s started,” Dr. Offner said.

Patients with traumatic brain injury are at increased risk for VTE because of multiple mechanisms, some related to injury and some to treatment. But efforts to prevent VTE in these patients are particularly challenging, because prophylaxis with an agent such as low-molecular-weight heparin also increases risk of intracranial hemorrhage, he said.

Dr. Offner and colleagues recently published a study suggesting that PTP use is associated with a 13-fold increase in odds of further hemorrhage progression in patients with unstable TBI, defined as evidence of intracranial hemorrhage progression on a follow-up CT scan within 1 day of admission (J. Trauma 2010;68:886-94).

To get a better handle on the optimal administration of PTP in patients with stable TBI and to see whether late or interrupted PTP might increase the risk of VTE, the investigators conducted a retrospective cohort study.

They reviewed records on all adult patients admitted to two urban level 1 trauma centers with a diagnosis of TBI based on ICD-9 codes. Patients with hospital stays shorter than 3 days, new or aggressive intracranial hemorrhage within 24 hours, or VTE diagnosed within 24 hours of admission were excluded.

The VTE prophylaxis protocol included placement of bilateral sequential compression devices (unless precluded by lower-extremity injuries) on admission to the ICU, and Lovenox (enoxaparin) 30 mg subcutaneously every 12 hours beginning at 36 hours, unless it was specifically held by neurosurgeons or other members of the care team. The patients also received weekly duplex ultrasound surveillance of the lower extremities.

The authors collected data on development of deep vein thrombosis (DVT) or pulmonary embolism (PE), DVT prophylaxis timing and method, and demographics. The analysis included incidence of VTE, associations between VTE and PTP – any exposure, early exposure (within 72 hours) compared with late, and interrupted administration of PTP – and multivariate logistic regression modeling.

Information on a total of 480 patients was available. The median age was 53 years, 62% were men, and 48% of injuries were due to falls. The mean Injury Severity Score was 19.7, mean Abbreviated Injury Scale (AIS) score was 3.5, and mean Glasgow Coma Scale (GCS) score was 12.2.

There were 15 venous thromboembolic events: 12 DVT and 5 PE (two patients had DVT and PE). The incidence was lower than the investigators had anticipated, Dr. Offner said.

In all, 93% of the patients were treated with sequential compression, and 255 received PTP. Of these, 42% received PTP early, 58% received it late, 74% received it continuously, and 26% had PTP interrupted.

Factors significantly related to the subsequent development of VTE included a GCS score of 3-8, compared with 9-15 for patients who did not develop a VTE; severe head injury (AIS score = 5); chest or extremity injury (AIS score 3 or greater for each); and less ambulation before discharge.

“When we looked at the univariate association between VTE and PTP, we were surprised in that administration of PTP either at all or in a late or early fashion didn’t seem to affect the subsequent incidence of VTE. On the other hand, if PTP was interrupted after it was started, there was a fourfold increase [odds ratio, 4.53; P = .02] in the incidence of subsequent venous thromboembolism,” Dr. Offner said.

In logistic regression analyses controlling for various risk factors, the investigators first compared PTP with no PTP and VTE incidence, and found that the only significant risk factor was ambulation before discharge (OR, 0.19; P = .02). Severity of injuries, preinjury anticoagulant/platelet inhibitor use, low GCS score, or female sex were not significantly linked to VTE incidence, regardless of PTP use.

There were also no significant differences in VTE incidence in late, compared with early administration of PTP.

However, when they looked at interrupted compared with continuous administration, they found that interrupted administration was associated with a sevenfold increased risk for VTE (OR, 7.07; P = .04). None of the other variables were significantly linked to VTE risk in this subanalysis.

Dr. Offner noted that the study results were limited by the retrospective design, potential for surveillance bias, and relatively small number of VTE cases in the sample.

Invited discussant Dr. Thomas Esposito of the department of surgery at Loyola University Medical Center in Maywood, Ill., noted that there was a mismatch between the interrupted and continuous PTP administration groups. That called the data into question, he said, because differences between the groups might have significantly influenced the incidence of VTE and PE.

Dr. Offner acknowledged that the small numbers involved could have subjected the study to a type II statistical error. He also agreed that the interrupted therapy group was a higher-risk group and that those risk factors could in part explain the higher incidence of VTE.

The study was internally funded. Neither Dr. Offner nor Dr. Esposito had conflict of interest disclosures.

BOSTON – Venous thromboembolism is a common complication of traumatic brain injury, and pharmacologic thromboprophylaxis is considered to be the standard of care. But when that prophylaxis is interrupted for surgery or other procedures, the patient’s risk for VTE significantly increases, reported investigators at the annual meeting of the American Association for the Surgery of Trauma.

Among 480 patients with stable traumatic brain injury (TBI), those who received interrupted pharmacologic thromboprophylaxis (PTP) had a sevenfold higher risk for VTE than did patients who received continuous prophylaxis, said Dr. Patrick Offner of the trauma service at St. Anthony Central Hospital in Denver.

“We would suggest that interruption of PTP in patients with traumatic brain injury does increase the risk of venous thromboembolism, and all efforts should be made to not interrupt it once it’s started,” Dr. Offner said.

Patients with traumatic brain injury are at increased risk for VTE because of multiple mechanisms, some related to injury and some to treatment. But efforts to prevent VTE in these patients are particularly challenging, because prophylaxis with an agent such as low-molecular-weight heparin also increases risk of intracranial hemorrhage, he said.

Dr. Offner and colleagues recently published a study suggesting that PTP use is associated with a 13-fold increase in odds of further hemorrhage progression in patients with unstable TBI, defined as evidence of intracranial hemorrhage progression on a follow-up CT scan within 1 day of admission (J. Trauma 2010;68:886-94).

To get a better handle on the optimal administration of PTP in patients with stable TBI and to see whether late or interrupted PTP might increase the risk of VTE, the investigators conducted a retrospective cohort study.

They reviewed records on all adult patients admitted to two urban level 1 trauma centers with a diagnosis of TBI based on ICD-9 codes. Patients with hospital stays shorter than 3 days, new or aggressive intracranial hemorrhage within 24 hours, or VTE diagnosed within 24 hours of admission were excluded.

The VTE prophylaxis protocol included placement of bilateral sequential compression devices (unless precluded by lower-extremity injuries) on admission to the ICU, and Lovenox (enoxaparin) 30 mg subcutaneously every 12 hours beginning at 36 hours, unless it was specifically held by neurosurgeons or other members of the care team. The patients also received weekly duplex ultrasound surveillance of the lower extremities.

The authors collected data on development of deep vein thrombosis (DVT) or pulmonary embolism (PE), DVT prophylaxis timing and method, and demographics. The analysis included incidence of VTE, associations between VTE and PTP – any exposure, early exposure (within 72 hours) compared with late, and interrupted administration of PTP – and multivariate logistic regression modeling.

Information on a total of 480 patients was available. The median age was 53 years, 62% were men, and 48% of injuries were due to falls. The mean Injury Severity Score was 19.7, mean Abbreviated Injury Scale (AIS) score was 3.5, and mean Glasgow Coma Scale (GCS) score was 12.2.

There were 15 venous thromboembolic events: 12 DVT and 5 PE (two patients had DVT and PE). The incidence was lower than the investigators had anticipated, Dr. Offner said.

In all, 93% of the patients were treated with sequential compression, and 255 received PTP. Of these, 42% received PTP early, 58% received it late, 74% received it continuously, and 26% had PTP interrupted.

Factors significantly related to the subsequent development of VTE included a GCS score of 3-8, compared with 9-15 for patients who did not develop a VTE; severe head injury (AIS score = 5); chest or extremity injury (AIS score 3 or greater for each); and less ambulation before discharge.

“When we looked at the univariate association between VTE and PTP, we were surprised in that administration of PTP either at all or in a late or early fashion didn’t seem to affect the subsequent incidence of VTE. On the other hand, if PTP was interrupted after it was started, there was a fourfold increase [odds ratio, 4.53; P = .02] in the incidence of subsequent venous thromboembolism,” Dr. Offner said.

In logistic regression analyses controlling for various risk factors, the investigators first compared PTP with no PTP and VTE incidence, and found that the only significant risk factor was ambulation before discharge (OR, 0.19; P = .02). Severity of injuries, preinjury anticoagulant/platelet inhibitor use, low GCS score, or female sex were not significantly linked to VTE incidence, regardless of PTP use.

There were also no significant differences in VTE incidence in late, compared with early administration of PTP.

However, when they looked at interrupted compared with continuous administration, they found that interrupted administration was associated with a sevenfold increased risk for VTE (OR, 7.07; P = .04). None of the other variables were significantly linked to VTE risk in this subanalysis.

Dr. Offner noted that the study results were limited by the retrospective design, potential for surveillance bias, and relatively small number of VTE cases in the sample.

Invited discussant Dr. Thomas Esposito of the department of surgery at Loyola University Medical Center in Maywood, Ill., noted that there was a mismatch between the interrupted and continuous PTP administration groups. That called the data into question, he said, because differences between the groups might have significantly influenced the incidence of VTE and PE.

Dr. Offner acknowledged that the small numbers involved could have subjected the study to a type II statistical error. He also agreed that the interrupted therapy group was a higher-risk group and that those risk factors could in part explain the higher incidence of VTE.

The study was internally funded. Neither Dr. Offner nor Dr. Esposito had conflict of interest disclosures.

BOSTON – Venous thromboembolism is a common complication of traumatic brain injury, and pharmacologic thromboprophylaxis is considered to be the standard of care. But when that prophylaxis is interrupted for surgery or other procedures, the patient’s risk for VTE significantly increases, reported investigators at the annual meeting of the American Association for the Surgery of Trauma.

Among 480 patients with stable traumatic brain injury (TBI), those who received interrupted pharmacologic thromboprophylaxis (PTP) had a sevenfold higher risk for VTE than did patients who received continuous prophylaxis, said Dr. Patrick Offner of the trauma service at St. Anthony Central Hospital in Denver.

“We would suggest that interruption of PTP in patients with traumatic brain injury does increase the risk of venous thromboembolism, and all efforts should be made to not interrupt it once it’s started,” Dr. Offner said.

Patients with traumatic brain injury are at increased risk for VTE because of multiple mechanisms, some related to injury and some to treatment. But efforts to prevent VTE in these patients are particularly challenging, because prophylaxis with an agent such as low-molecular-weight heparin also increases risk of intracranial hemorrhage, he said.

Dr. Offner and colleagues recently published a study suggesting that PTP use is associated with a 13-fold increase in odds of further hemorrhage progression in patients with unstable TBI, defined as evidence of intracranial hemorrhage progression on a follow-up CT scan within 1 day of admission (J. Trauma 2010;68:886-94).

To get a better handle on the optimal administration of PTP in patients with stable TBI and to see whether late or interrupted PTP might increase the risk of VTE, the investigators conducted a retrospective cohort study.

They reviewed records on all adult patients admitted to two urban level 1 trauma centers with a diagnosis of TBI based on ICD-9 codes. Patients with hospital stays shorter than 3 days, new or aggressive intracranial hemorrhage within 24 hours, or VTE diagnosed within 24 hours of admission were excluded.

The VTE prophylaxis protocol included placement of bilateral sequential compression devices (unless precluded by lower-extremity injuries) on admission to the ICU, and Lovenox (enoxaparin) 30 mg subcutaneously every 12 hours beginning at 36 hours, unless it was specifically held by neurosurgeons or other members of the care team. The patients also received weekly duplex ultrasound surveillance of the lower extremities.

The authors collected data on development of deep vein thrombosis (DVT) or pulmonary embolism (PE), DVT prophylaxis timing and method, and demographics. The analysis included incidence of VTE, associations between VTE and PTP – any exposure, early exposure (within 72 hours) compared with late, and interrupted administration of PTP – and multivariate logistic regression modeling.

Information on a total of 480 patients was available. The median age was 53 years, 62% were men, and 48% of injuries were due to falls. The mean Injury Severity Score was 19.7, mean Abbreviated Injury Scale (AIS) score was 3.5, and mean Glasgow Coma Scale (GCS) score was 12.2.

There were 15 venous thromboembolic events: 12 DVT and 5 PE (two patients had DVT and PE). The incidence was lower than the investigators had anticipated, Dr. Offner said.

In all, 93% of the patients were treated with sequential compression, and 255 received PTP. Of these, 42% received PTP early, 58% received it late, 74% received it continuously, and 26% had PTP interrupted.

Factors significantly related to the subsequent development of VTE included a GCS score of 3-8, compared with 9-15 for patients who did not develop a VTE; severe head injury (AIS score = 5); chest or extremity injury (AIS score 3 or greater for each); and less ambulation before discharge.

“When we looked at the univariate association between VTE and PTP, we were surprised in that administration of PTP either at all or in a late or early fashion didn’t seem to affect the subsequent incidence of VTE. On the other hand, if PTP was interrupted after it was started, there was a fourfold increase [odds ratio, 4.53; P = .02] in the incidence of subsequent venous thromboembolism,” Dr. Offner said.

In logistic regression analyses controlling for various risk factors, the investigators first compared PTP with no PTP and VTE incidence, and found that the only significant risk factor was ambulation before discharge (OR, 0.19; P = .02). Severity of injuries, preinjury anticoagulant/platelet inhibitor use, low GCS score, or female sex were not significantly linked to VTE incidence, regardless of PTP use.

There were also no significant differences in VTE incidence in late, compared with early administration of PTP.

However, when they looked at interrupted compared with continuous administration, they found that interrupted administration was associated with a sevenfold increased risk for VTE (OR, 7.07; P = .04). None of the other variables were significantly linked to VTE risk in this subanalysis.

Dr. Offner noted that the study results were limited by the retrospective design, potential for surveillance bias, and relatively small number of VTE cases in the sample.

Invited discussant Dr. Thomas Esposito of the department of surgery at Loyola University Medical Center in Maywood, Ill., noted that there was a mismatch between the interrupted and continuous PTP administration groups. That called the data into question, he said, because differences between the groups might have significantly influenced the incidence of VTE and PE.

Dr. Offner acknowledged that the small numbers involved could have subjected the study to a type II statistical error. He also agreed that the interrupted therapy group was a higher-risk group and that those risk factors could in part explain the higher incidence of VTE.

The study was internally funded. Neither Dr. Offner nor Dr. Esposito had conflict of interest disclosures.

BOSTON – Venous thromboembolism is a common complication of traumatic brain injury, and pharmacologic thromboprophylaxis is considered to be the standard of care. But when that prophylaxis is interrupted for surgery or other procedures, the patient’s risk for VTE significantly increases, reported investigators at the annual meeting of the American Association for the Surgery of Trauma.

Among 480 patients with stable traumatic brain injury (TBI), those who received interrupted pharmacologic thromboprophylaxis (PTP) had a sevenfold higher risk for VTE than did patients who received continuous prophylaxis, said Dr. Patrick Offner of the trauma service at St. Anthony Central Hospital in Denver.

“We would suggest that interruption of PTP in patients with traumatic brain injury does increase the risk of venous thromboembolism, and all efforts should be made to not interrupt it once it’s started,” Dr. Offner said.

Patients with traumatic brain injury are at increased risk for VTE because of multiple mechanisms, some related to injury and some to treatment. But efforts to prevent VTE in these patients are particularly challenging, because prophylaxis with an agent such as low-molecular-weight heparin also increases risk of intracranial hemorrhage, he said.

Dr. Offner and colleagues recently published a study suggesting that PTP use is associated with a 13-fold increase in odds of further hemorrhage progression in patients with unstable TBI, defined as evidence of intracranial hemorrhage progression on a follow-up CT scan within 1 day of admission (J. Trauma 2010;68:886-94).

To get a better handle on the optimal administration of PTP in patients with stable TBI and to see whether late or interrupted PTP might increase the risk of VTE, the investigators conducted a retrospective cohort study.

They reviewed records on all adult patients admitted to two urban level 1 trauma centers with a diagnosis of TBI based on ICD-9 codes. Patients with hospital stays shorter than 3 days, new or aggressive intracranial hemorrhage within 24 hours, or VTE diagnosed within 24 hours of admission were excluded.

The VTE prophylaxis protocol included placement of bilateral sequential compression devices (unless precluded by lower-extremity injuries) on admission to the ICU, and Lovenox (enoxaparin) 30 mg subcutaneously every 12 hours beginning at 36 hours, unless it was specifically held by neurosurgeons or other members of the care team. The patients also received weekly duplex ultrasound surveillance of the lower extremities.

The authors collected data on development of deep vein thrombosis (DVT) or pulmonary embolism (PE), DVT prophylaxis timing and method, and demographics. The analysis included incidence of VTE, associations between VTE and PTP – any exposure, early exposure (within 72 hours) compared with late, and interrupted administration of PTP – and multivariate logistic regression modeling.

Information on a total of 480 patients was available. The median age was 53 years, 62% were men, and 48% of injuries were due to falls. The mean Injury Severity Score was 19.7, mean Abbreviated Injury Scale (AIS) score was 3.5, and mean Glasgow Coma Scale (GCS) score was 12.2.

There were 15 venous thromboembolic events: 12 DVT and 5 PE (two patients had DVT and PE). The incidence was lower than the investigators had anticipated, Dr. Offner said.

In all, 93% of the patients were treated with sequential compression, and 255 received PTP. Of these, 42% received PTP early, 58% received it late, 74% received it continuously, and 26% had PTP interrupted.

Factors significantly related to the subsequent development of VTE included a GCS score of 3-8, compared with 9-15 for patients who did not develop a VTE; severe head injury (AIS score = 5); chest or extremity injury (AIS score 3 or greater for each); and less ambulation before discharge.

“When we looked at the univariate association between VTE and PTP, we were surprised in that administration of PTP either at all or in a late or early fashion didn’t seem to affect the subsequent incidence of VTE. On the other hand, if PTP was interrupted after it was started, there was a fourfold increase [odds ratio, 4.53; P = .02] in the incidence of subsequent venous thromboembolism,” Dr. Offner said.

In logistic regression analyses controlling for various risk factors, the investigators first compared PTP with no PTP and VTE incidence, and found that the only significant risk factor was ambulation before discharge (OR, 0.19; P = .02). Severity of injuries, preinjury anticoagulant/platelet inhibitor use, low GCS score, or female sex were not significantly linked to VTE incidence, regardless of PTP use.

There were also no significant differences in VTE incidence in late, compared with early administration of PTP.

However, when they looked at interrupted compared with continuous administration, they found that interrupted administration was associated with a sevenfold increased risk for VTE (OR, 7.07; P = .04). None of the other variables were significantly linked to VTE risk in this subanalysis.

Dr. Offner noted that the study results were limited by the retrospective design, potential for surveillance bias, and relatively small number of VTE cases in the sample.

Invited discussant Dr. Thomas Esposito of the department of surgery at Loyola University Medical Center in Maywood, Ill., noted that there was a mismatch between the interrupted and continuous PTP administration groups. That called the data into question, he said, because differences between the groups might have significantly influenced the incidence of VTE and PE.

Dr. Offner acknowledged that the small numbers involved could have subjected the study to a type II statistical error. He also agreed that the interrupted therapy group was a higher-risk group and that those risk factors could in part explain the higher incidence of VTE.

The study was internally funded. Neither Dr. Offner nor Dr. Esposito had conflict of interest disclosures.

BOSTON – Venous thromboembolism is a common complication of traumatic brain injury, and pharmacologic thromboprophylaxis is considered to be the standard of care. But when that prophylaxis is interrupted for surgery or other procedures, the patient’s risk for VTE significantly increases, reported investigators at the annual meeting of the American Association for the Surgery of Trauma.

Among 480 patients with stable traumatic brain injury (TBI), those who received interrupted pharmacologic thromboprophylaxis (PTP) had a sevenfold higher risk for VTE than did patients who received continuous prophylaxis, said Dr. Patrick Offner of the trauma service at St. Anthony Central Hospital in Denver.

“We would suggest that interruption of PTP in patients with traumatic brain injury does increase the risk of venous thromboembolism, and all efforts should be made to not interrupt it once it’s started,” Dr. Offner said.

Patients with traumatic brain injury are at increased risk for VTE because of multiple mechanisms, some related to injury and some to treatment. But efforts to prevent VTE in these patients are particularly challenging, because prophylaxis with an agent such as low-molecular-weight heparin also increases risk of intracranial hemorrhage, he said.

Dr. Offner and colleagues recently published a study suggesting that PTP use is associated with a 13-fold increase in odds of further hemorrhage progression in patients with unstable TBI, defined as evidence of intracranial hemorrhage progression on a follow-up CT scan within 1 day of admission (J. Trauma 2010;68:886-94).

To get a better handle on the optimal administration of PTP in patients with stable TBI and to see whether late or interrupted PTP might increase the risk of VTE, the investigators conducted a retrospective cohort study.

They reviewed records on all adult patients admitted to two urban level 1 trauma centers with a diagnosis of TBI based on ICD-9 codes. Patients with hospital stays shorter than 3 days, new or aggressive intracranial hemorrhage within 24 hours, or VTE diagnosed within 24 hours of admission were excluded.

The VTE prophylaxis protocol included placement of bilateral sequential compression devices (unless precluded by lower-extremity injuries) on admission to the ICU, and Lovenox (enoxaparin) 30 mg subcutaneously every 12 hours beginning at 36 hours, unless it was specifically held by neurosurgeons or other members of the care team. The patients also received weekly duplex ultrasound surveillance of the lower extremities.

The authors collected data on development of deep vein thrombosis (DVT) or pulmonary embolism (PE), DVT prophylaxis timing and method, and demographics. The analysis included incidence of VTE, associations between VTE and PTP – any exposure, early exposure (within 72 hours) compared with late, and interrupted administration of PTP – and multivariate logistic regression modeling.

Information on a total of 480 patients was available. The median age was 53 years, 62% were men, and 48% of injuries were due to falls. The mean Injury Severity Score was 19.7, mean Abbreviated Injury Scale (AIS) score was 3.5, and mean Glasgow Coma Scale (GCS) score was 12.2.

There were 15 venous thromboembolic events: 12 DVT and 5 PE (two patients had DVT and PE). The incidence was lower than the investigators had anticipated, Dr. Offner said.

In all, 93% of the patients were treated with sequential compression, and 255 received PTP. Of these, 42% received PTP early, 58% received it late, 74% received it continuously, and 26% had PTP interrupted.

Factors significantly related to the subsequent development of VTE included a GCS score of 3-8, compared with 9-15 for patients who did not develop a VTE; severe head injury (AIS score = 5); chest or extremity injury (AIS score 3 or greater for each); and less ambulation before discharge.

“When we looked at the univariate association between VTE and PTP, we were surprised in that administration of PTP either at all or in a late or early fashion didn’t seem to affect the subsequent incidence of VTE. On the other hand, if PTP was interrupted after it was started, there was a fourfold increase [odds ratio, 4.53; P = .02] in the incidence of subsequent venous thromboembolism,” Dr. Offner said.

In logistic regression analyses controlling for various risk factors, the investigators first compared PTP with no PTP and VTE incidence, and found that the only significant risk factor was ambulation before discharge (OR, 0.19; P = .02). Severity of injuries, preinjury anticoagulant/platelet inhibitor use, low GCS score, or female sex were not significantly linked to VTE incidence, regardless of PTP use.

There were also no significant differences in VTE incidence in late, compared with early administration of PTP.

However, when they looked at interrupted compared with continuous administration, they found that interrupted administration was associated with a sevenfold increased risk for VTE (OR, 7.07; P = .04). None of the other variables were significantly linked to VTE risk in this subanalysis.

Dr. Offner noted that the study results were limited by the retrospective design, potential for surveillance bias, and relatively small number of VTE cases in the sample.

Invited discussant Dr. Thomas Esposito of the department of surgery at Loyola University Medical Center in Maywood, Ill., noted that there was a mismatch between the interrupted and continuous PTP administration groups. That called the data into question, he said, because differences between the groups might have significantly influenced the incidence of VTE and PE.

Dr. Offner acknowledged that the small numbers involved could have subjected the study to a type II statistical error. He also agreed that the interrupted therapy group was a higher-risk group and that those risk factors could in part explain the higher incidence of VTE.

The study was internally funded. Neither Dr. Offner nor Dr. Esposito had conflict of interest disclosures.

BOSTON – Trauma patients with low levels of ionized calcium at admission were at increased risk for death and coagulopathies and were more likely to need massive transfusions, compared with patients with normal physiologic levels.

Among 694 consecutive trauma activations over an 18-month period, the death rate for patients with ionized calcium levels below 1.0 mmol/L was 21%, compared with 13% for patients with levels of 1.0 mmol/L or higher (P = .016), Dr. Ann P. O’Rourke, of the division of general surgery at the University of Wisconsin in Madison, said at the annual meeting of the American Association for the Surgery of Trauma.

Ionized calcium levels also remained an independent predictor for massive transfusion requirements (10 or more units of packed red blood cells over 24 hours) in a logistic regression analysis controlling for age, sex, coagulopathy, and severity of injury (odds ratio, 2.557; P = .0004).

“Metabolic and cellular derangements related to shock occur rapidly post injury. We need to continue to identify reliable and reproducible markers of outcome, and ionized calcium levels drawn prior to any resuscitative efforts may serve as a reasonable marker,” Dr. O’Rourke said.

Calcium is a ubiquitous component of the clotting cascade, and is necessary for platelet activation and thrombus formation, she noted.

“Hypocalcemia has been shown to be a common metabolic derangement in critically ill patients. It has been shown to be prevalent in patients who are septic, in patients who have complications related to musculoskeletal trauma, and in burn patients,” she said.

Previous studies have shown that low ionized calcium levels are associated with hypotension and predict mortality, she added.

To see whether, as they hypothesized, low levels of ionized calcium at admission could predict massive transfusion, Dr. O’Rourke and colleagues at the University of Tennessee Health Science Center in Memphis, where she served a trauma surgery fellowship, prospectively studied all 694 patients with trauma initially treated at the center over an 18-month period. Patients transferred from other centers were excluded.

They collected data on demographics, injury mechanism, vital signs, 24-hour transfusion requirements, and mortality, and compared outcomes using Wilcoxon rank-sum and chi-square tests.

They defined low calcium as a level of 1.0 mmol/L or less, according to Youden’s index, which determines optimal cut-points as the area on a receiver operating characteristic curve where the highest degrees of test sensitivity and specificity meet.

The patients had a mean age of 38 years, 77% were men, and 67% had blunt trauma injuries. The mean Glasgow Coma Scale (GCS) score was 12, and the mean Injury Severity Score (ISS) was 23.

In all, 291 patients had high levels of ionized calcium (1.0 mmol/L or greater) and 403 had low levels. The high and low groups were evenly matched by age, sex, injury mechanism, and lactate levels, but GCS scores were significantly higher in the high-calcium group (12 vs. 11, respectively; P = .020), whereas ISS was higher in the low-calcium group (17 vs. 23; P = .003). Base deficit, a measure of the severity of shock, was lower in the low-calcium patients (–4.7 vs. –2.9; P less than .0001).

In addition to the higher mortality among patients low in calcium, these patients had significantly higher levels of coagulopathy, as measured by a mean international normalized ratio (INR) of 1.46 on arrival in the trauma bay, compared with 1.19 for patients with high calcium levels. Massive transfusions were required in 29% of the patients with low levels of ionized calcium, compared with 12% of the group with high levels (P less than .0001).

A multivariable logistic regression analysis showed that low ionized calcium level was the only significant predictor of transfusion. Neither age, admission GCS, ISS, or coagulopathy significantly predicted transfusion requirements.

“Admission ionized calcium levels may facilitate rapid identification of patients requiring massive transfusion, allow us to have earlier preparation and administration of blood products, and may serve as a trigger to our blood bank to initiate our massive transfusion protocols,” Dr. O’Rourke said.

Dr. Matthew Rosengart, the invited discussant, noted that “for decades we have recognized that alterations in calcium, such as hypocalcemia, develop in our sickest of patients with prevalence upwards of 88% in the presence of severe sepsis. The elusive question is why.”

Further investigations should focus on determining how ionized calcium levels fit in with other scoring instruments that incorporate immediately available biochemical parameters, he said.

“How does this parameter, calcium, perform when other validated scores are incorporated into your model? Does it retain its association after more rigorous adjustment for case mix, including base deficit or lactate? Should ionized calcium complement or replace these instruments?” asked Dr. Rosengart, of the University of Pittsburgh Medical Center.

“I would say that any additional tool we can have to make something more predictive, so that our decision making is better, is useful,” Dr. O’Rourke replied.

The funding source for the study was not specified. The authors and Dr. Rosengart had no conflict of interest disclosures.

BOSTON – Trauma patients with low levels of ionized calcium at admission were at increased risk for death and coagulopathies and were more likely to need massive transfusions, compared with patients with normal physiologic levels.

Among 694 consecutive trauma activations over an 18-month period, the death rate for patients with ionized calcium levels below 1.0 mmol/L was 21%, compared with 13% for patients with levels of 1.0 mmol/L or higher (P = .016), Dr. Ann P. O’Rourke, of the division of general surgery at the University of Wisconsin in Madison, said at the annual meeting of the American Association for the Surgery of Trauma.

Ionized calcium levels also remained an independent predictor for massive transfusion requirements (10 or more units of packed red blood cells over 24 hours) in a logistic regression analysis controlling for age, sex, coagulopathy, and severity of injury (odds ratio, 2.557; P = .0004).

“Metabolic and cellular derangements related to shock occur rapidly post injury. We need to continue to identify reliable and reproducible markers of outcome, and ionized calcium levels drawn prior to any resuscitative efforts may serve as a reasonable marker,” Dr. O’Rourke said.

Calcium is a ubiquitous component of the clotting cascade, and is necessary for platelet activation and thrombus formation, she noted.

“Hypocalcemia has been shown to be a common metabolic derangement in critically ill patients. It has been shown to be prevalent in patients who are septic, in patients who have complications related to musculoskeletal trauma, and in burn patients,” she said.

Previous studies have shown that low ionized calcium levels are associated with hypotension and predict mortality, she added.

To see whether, as they hypothesized, low levels of ionized calcium at admission could predict massive transfusion, Dr. O’Rourke and colleagues at the University of Tennessee Health Science Center in Memphis, where she served a trauma surgery fellowship, prospectively studied all 694 patients with trauma initially treated at the center over an 18-month period. Patients transferred from other centers were excluded.

They collected data on demographics, injury mechanism, vital signs, 24-hour transfusion requirements, and mortality, and compared outcomes using Wilcoxon rank-sum and chi-square tests.

They defined low calcium as a level of 1.0 mmol/L or less, according to Youden’s index, which determines optimal cut-points as the area on a receiver operating characteristic curve where the highest degrees of test sensitivity and specificity meet.

The patients had a mean age of 38 years, 77% were men, and 67% had blunt trauma injuries. The mean Glasgow Coma Scale (GCS) score was 12, and the mean Injury Severity Score (ISS) was 23.

In all, 291 patients had high levels of ionized calcium (1.0 mmol/L or greater) and 403 had low levels. The high and low groups were evenly matched by age, sex, injury mechanism, and lactate levels, but GCS scores were significantly higher in the high-calcium group (12 vs. 11, respectively; P = .020), whereas ISS was higher in the low-calcium group (17 vs. 23; P = .003). Base deficit, a measure of the severity of shock, was lower in the low-calcium patients (–4.7 vs. –2.9; P less than .0001).

In addition to the higher mortality among patients low in calcium, these patients had significantly higher levels of coagulopathy, as measured by a mean international normalized ratio (INR) of 1.46 on arrival in the trauma bay, compared with 1.19 for patients with high calcium levels. Massive transfusions were required in 29% of the patients with low levels of ionized calcium, compared with 12% of the group with high levels (P less than .0001).

A multivariable logistic regression analysis showed that low ionized calcium level was the only significant predictor of transfusion. Neither age, admission GCS, ISS, or coagulopathy significantly predicted transfusion requirements.

“Admission ionized calcium levels may facilitate rapid identification of patients requiring massive transfusion, allow us to have earlier preparation and administration of blood products, and may serve as a trigger to our blood bank to initiate our massive transfusion protocols,” Dr. O’Rourke said.

Dr. Matthew Rosengart, the invited discussant, noted that “for decades we have recognized that alterations in calcium, such as hypocalcemia, develop in our sickest of patients with prevalence upwards of 88% in the presence of severe sepsis. The elusive question is why.”

Further investigations should focus on determining how ionized calcium levels fit in with other scoring instruments that incorporate immediately available biochemical parameters, he said.

“How does this parameter, calcium, perform when other validated scores are incorporated into your model? Does it retain its association after more rigorous adjustment for case mix, including base deficit or lactate? Should ionized calcium complement or replace these instruments?” asked Dr. Rosengart, of the University of Pittsburgh Medical Center.

“I would say that any additional tool we can have to make something more predictive, so that our decision making is better, is useful,” Dr. O’Rourke replied.

The funding source for the study was not specified. The authors and Dr. Rosengart had no conflict of interest disclosures.

BOSTON – Trauma patients with low levels of ionized calcium at admission were at increased risk for death and coagulopathies and were more likely to need massive transfusions, compared with patients with normal physiologic levels.

Among 694 consecutive trauma activations over an 18-month period, the death rate for patients with ionized calcium levels below 1.0 mmol/L was 21%, compared with 13% for patients with levels of 1.0 mmol/L or higher (P = .016), Dr. Ann P. O’Rourke, of the division of general surgery at the University of Wisconsin in Madison, said at the annual meeting of the American Association for the Surgery of Trauma.

Ionized calcium levels also remained an independent predictor for massive transfusion requirements (10 or more units of packed red blood cells over 24 hours) in a logistic regression analysis controlling for age, sex, coagulopathy, and severity of injury (odds ratio, 2.557; P = .0004).

“Metabolic and cellular derangements related to shock occur rapidly post injury. We need to continue to identify reliable and reproducible markers of outcome, and ionized calcium levels drawn prior to any resuscitative efforts may serve as a reasonable marker,” Dr. O’Rourke said.

Calcium is a ubiquitous component of the clotting cascade, and is necessary for platelet activation and thrombus formation, she noted.

“Hypocalcemia has been shown to be a common metabolic derangement in critically ill patients. It has been shown to be prevalent in patients who are septic, in patients who have complications related to musculoskeletal trauma, and in burn patients,” she said.

Previous studies have shown that low ionized calcium levels are associated with hypotension and predict mortality, she added.

To see whether, as they hypothesized, low levels of ionized calcium at admission could predict massive transfusion, Dr. O’Rourke and colleagues at the University of Tennessee Health Science Center in Memphis, where she served a trauma surgery fellowship, prospectively studied all 694 patients with trauma initially treated at the center over an 18-month period. Patients transferred from other centers were excluded.

They collected data on demographics, injury mechanism, vital signs, 24-hour transfusion requirements, and mortality, and compared outcomes using Wilcoxon rank-sum and chi-square tests.

They defined low calcium as a level of 1.0 mmol/L or less, according to Youden’s index, which determines optimal cut-points as the area on a receiver operating characteristic curve where the highest degrees of test sensitivity and specificity meet.

The patients had a mean age of 38 years, 77% were men, and 67% had blunt trauma injuries. The mean Glasgow Coma Scale (GCS) score was 12, and the mean Injury Severity Score (ISS) was 23.

In all, 291 patients had high levels of ionized calcium (1.0 mmol/L or greater) and 403 had low levels. The high and low groups were evenly matched by age, sex, injury mechanism, and lactate levels, but GCS scores were significantly higher in the high-calcium group (12 vs. 11, respectively; P = .020), whereas ISS was higher in the low-calcium group (17 vs. 23; P = .003). Base deficit, a measure of the severity of shock, was lower in the low-calcium patients (–4.7 vs. –2.9; P less than .0001).

In addition to the higher mortality among patients low in calcium, these patients had significantly higher levels of coagulopathy, as measured by a mean international normalized ratio (INR) of 1.46 on arrival in the trauma bay, compared with 1.19 for patients with high calcium levels. Massive transfusions were required in 29% of the patients with low levels of ionized calcium, compared with 12% of the group with high levels (P less than .0001).

A multivariable logistic regression analysis showed that low ionized calcium level was the only significant predictor of transfusion. Neither age, admission GCS, ISS, or coagulopathy significantly predicted transfusion requirements.

“Admission ionized calcium levels may facilitate rapid identification of patients requiring massive transfusion, allow us to have earlier preparation and administration of blood products, and may serve as a trigger to our blood bank to initiate our massive transfusion protocols,” Dr. O’Rourke said.

Dr. Matthew Rosengart, the invited discussant, noted that “for decades we have recognized that alterations in calcium, such as hypocalcemia, develop in our sickest of patients with prevalence upwards of 88% in the presence of severe sepsis. The elusive question is why.”

Further investigations should focus on determining how ionized calcium levels fit in with other scoring instruments that incorporate immediately available biochemical parameters, he said.

“How does this parameter, calcium, perform when other validated scores are incorporated into your model? Does it retain its association after more rigorous adjustment for case mix, including base deficit or lactate? Should ionized calcium complement or replace these instruments?” asked Dr. Rosengart, of the University of Pittsburgh Medical Center.

“I would say that any additional tool we can have to make something more predictive, so that our decision making is better, is useful,” Dr. O’Rourke replied.

The funding source for the study was not specified. The authors and Dr. Rosengart had no conflict of interest disclosures.

BOSTON – Hyperchloremia in patients with Clostridium difficile infections is an indicator of disease severity and a risk factor for death within 30 days, according to Dr. Anilrudh A. Venugopal.

A retrospective chart review of 136 patients with C. difficile infections treated over an 8-month period showed that hyperchloremia was the only electrolyte abnormality associated with severe C. difficile infection, as determined by using two C. difficile infection scoring systems, Dr. Venugopal said at the conference.

Among patients not on hemodialysis, hyperchloremia was also associated with an increased risk of all-cause mortality at 30 days, based on a multivariate analysis. Other significant predictors of death were mean albumin and creatinine levels, and the presence of malignant solid tumors.

“By identifying patients with Clostridium difficile infection that are at risk for severe disease, it may affect our initial management,” said Dr. Venugopal of the division of infectious disease at the St. John Hospital and Medical Center in Grosse Pointe Woods, Mich.

Because severe diarrheal disease can result in a metabolic acidosis that is often hyperchloremic and hypokalemic, he and his coinvestigators sought to evaluate whether electrolyte abnormalities at diagnosis could predict severe disease and death.

They retrospectively studied charts of 58 men and 78 women treated from October 2009 through May 2010. The mean patient age was 67 years. In all, 91% of the patients had been admitted to a health care facility within 30 days of the initial diarrheal episode, and 88% had received antibiotics within 39 days. The all-cause mortality rate was 23% (31 deaths).

The most common comorbidities were coronary artery disease/heart failure in 47%, diabetes in 37%, chronic obstructive pulmonary disease in 22%, and moderate to severe renal disease in 19% (26 patients, 12 of whom required dialysis).

The investigators defined the severity of C. difficile infections by scores on the C. difficile infection scoring systems developed by Dr. Jaime Belmares (J. Infect. 2007;55:495-501) and Dr. Fred A. Zar (Clin. Infect. Dis. 2007;45:302-7).

In the initial univariate analysis, the authors sought to validate the scoring systems and found an association between death and a score of 2 or higher on each system, as well as a trend toward death and higher chloride levels.

To see whether conditions that cause electrolyte abnormalities, such as end-stage renal disease, could affect the association between chloride and mortality, the investigators evaluated lab results for the 12 patients on dialysis.

“What we found was that the dialysis patients had a significantly lower sodium and lower chloride level compared to the nondialysis population, and they had a higher potassium level,” Dr. Venugopal said.

The authors then removed the dialysis population and performed univariate analyses on 124 patients, which showed significant associations between death within 30 days and the mean chloride level (P = .04) and bicarbonate level (P = .05), but not the sodium or potassium levels. Other significant mortality risk factors in this group were agent change during therapy (P = .001), ICU stay prior to diagnosis (P = .006), solid tumor malignancies (P less than .0001), and HIV (P = .013).

In a univariate analysis stratified by scoring system, chloride alone among the electrolytes was significantly associated with a high severity score (P = .004 for the Belmares system; P less than .0001 for the Zar system).

In multivariate logistic regression analysis, again with only nondialysis patients, the authors found that the mean chloride level remained a significant predictor of 30-day mortality, with an odds ratio of 1.14 (P = .043). Other significant predictors in this model were solid tumors (OR, 14.08; P = .005), mean albumin level (OR, 0.91; P = .006), and mean creatinine level (OR, 2.12; P = .045). The study was internally funded. Neither Dr. Venugopal nor his coinvestigators said they had financial conflicts.

BOSTON – Hyperchloremia in patients with Clostridium difficile infections is an indicator of disease severity and a risk factor for death within 30 days, according to Dr. Anilrudh A. Venugopal.

A retrospective chart review of 136 patients with C. difficile infections treated over an 8-month period showed that hyperchloremia was the only electrolyte abnormality associated with severe C. difficile infection, as determined by using two C. difficile infection scoring systems, Dr. Venugopal said at the conference.

Among patients not on hemodialysis, hyperchloremia was also associated with an increased risk of all-cause mortality at 30 days, based on a multivariate analysis. Other significant predictors of death were mean albumin and creatinine levels, and the presence of malignant solid tumors.

“By identifying patients with Clostridium difficile infection that are at risk for severe disease, it may affect our initial management,” said Dr. Venugopal of the division of infectious disease at the St. John Hospital and Medical Center in Grosse Pointe Woods, Mich.

Because severe diarrheal disease can result in a metabolic acidosis that is often hyperchloremic and hypokalemic, he and his coinvestigators sought to evaluate whether electrolyte abnormalities at diagnosis could predict severe disease and death.

They retrospectively studied charts of 58 men and 78 women treated from October 2009 through May 2010. The mean patient age was 67 years. In all, 91% of the patients had been admitted to a health care facility within 30 days of the initial diarrheal episode, and 88% had received antibiotics within 39 days. The all-cause mortality rate was 23% (31 deaths).

The most common comorbidities were coronary artery disease/heart failure in 47%, diabetes in 37%, chronic obstructive pulmonary disease in 22%, and moderate to severe renal disease in 19% (26 patients, 12 of whom required dialysis).

The investigators defined the severity of C. difficile infections by scores on the C. difficile infection scoring systems developed by Dr. Jaime Belmares (J. Infect. 2007;55:495-501) and Dr. Fred A. Zar (Clin. Infect. Dis. 2007;45:302-7).

In the initial univariate analysis, the authors sought to validate the scoring systems and found an association between death and a score of 2 or higher on each system, as well as a trend toward death and higher chloride levels.

To see whether conditions that cause electrolyte abnormalities, such as end-stage renal disease, could affect the association between chloride and mortality, the investigators evaluated lab results for the 12 patients on dialysis.

“What we found was that the dialysis patients had a significantly lower sodium and lower chloride level compared to the nondialysis population, and they had a higher potassium level,” Dr. Venugopal said.

The authors then removed the dialysis population and performed univariate analyses on 124 patients, which showed significant associations between death within 30 days and the mean chloride level (P = .04) and bicarbonate level (P = .05), but not the sodium or potassium levels. Other significant mortality risk factors in this group were agent change during therapy (P = .001), ICU stay prior to diagnosis (P = .006), solid tumor malignancies (P less than .0001), and HIV (P = .013).

In a univariate analysis stratified by scoring system, chloride alone among the electrolytes was significantly associated with a high severity score (P = .004 for the Belmares system; P less than .0001 for the Zar system).

In multivariate logistic regression analysis, again with only nondialysis patients, the authors found that the mean chloride level remained a significant predictor of 30-day mortality, with an odds ratio of 1.14 (P = .043). Other significant predictors in this model were solid tumors (OR, 14.08; P = .005), mean albumin level (OR, 0.91; P = .006), and mean creatinine level (OR, 2.12; P = .045). The study was internally funded. Neither Dr. Venugopal nor his coinvestigators said they had financial conflicts.

BOSTON – Hyperchloremia in patients with Clostridium difficile infections is an indicator of disease severity and a risk factor for death within 30 days, according to Dr. Anilrudh A. Venugopal.

A retrospective chart review of 136 patients with C. difficile infections treated over an 8-month period showed that hyperchloremia was the only electrolyte abnormality associated with severe C. difficile infection, as determined by using two C. difficile infection scoring systems, Dr. Venugopal said at the conference.

Among patients not on hemodialysis, hyperchloremia was also associated with an increased risk of all-cause mortality at 30 days, based on a multivariate analysis. Other significant predictors of death were mean albumin and creatinine levels, and the presence of malignant solid tumors.

“By identifying patients with Clostridium difficile infection that are at risk for severe disease, it may affect our initial management,” said Dr. Venugopal of the division of infectious disease at the St. John Hospital and Medical Center in Grosse Pointe Woods, Mich.

Because severe diarrheal disease can result in a metabolic acidosis that is often hyperchloremic and hypokalemic, he and his coinvestigators sought to evaluate whether electrolyte abnormalities at diagnosis could predict severe disease and death.

They retrospectively studied charts of 58 men and 78 women treated from October 2009 through May 2010. The mean patient age was 67 years. In all, 91% of the patients had been admitted to a health care facility within 30 days of the initial diarrheal episode, and 88% had received antibiotics within 39 days. The all-cause mortality rate was 23% (31 deaths).

The most common comorbidities were coronary artery disease/heart failure in 47%, diabetes in 37%, chronic obstructive pulmonary disease in 22%, and moderate to severe renal disease in 19% (26 patients, 12 of whom required dialysis).

The investigators defined the severity of C. difficile infections by scores on the C. difficile infection scoring systems developed by Dr. Jaime Belmares (J. Infect. 2007;55:495-501) and Dr. Fred A. Zar (Clin. Infect. Dis. 2007;45:302-7).

In the initial univariate analysis, the authors sought to validate the scoring systems and found an association between death and a score of 2 or higher on each system, as well as a trend toward death and higher chloride levels.

To see whether conditions that cause electrolyte abnormalities, such as end-stage renal disease, could affect the association between chloride and mortality, the investigators evaluated lab results for the 12 patients on dialysis.

“What we found was that the dialysis patients had a significantly lower sodium and lower chloride level compared to the nondialysis population, and they had a higher potassium level,” Dr. Venugopal said.

The authors then removed the dialysis population and performed univariate analyses on 124 patients, which showed significant associations between death within 30 days and the mean chloride level (P = .04) and bicarbonate level (P = .05), but not the sodium or potassium levels. Other significant mortality risk factors in this group were agent change during therapy (P = .001), ICU stay prior to diagnosis (P = .006), solid tumor malignancies (P less than .0001), and HIV (P = .013).

In a univariate analysis stratified by scoring system, chloride alone among the electrolytes was significantly associated with a high severity score (P = .004 for the Belmares system; P less than .0001 for the Zar system).

In multivariate logistic regression analysis, again with only nondialysis patients, the authors found that the mean chloride level remained a significant predictor of 30-day mortality, with an odds ratio of 1.14 (P = .043). Other significant predictors in this model were solid tumors (OR, 14.08; P = .005), mean albumin level (OR, 0.91; P = .006), and mean creatinine level (OR, 2.12; P = .045). The study was internally funded. Neither Dr. Venugopal nor his coinvestigators said they had financial conflicts.

BOSTON – It seems counterintuitive, but when management times of potential heart and lung donors stretched beyond 20 hours, successful organ procurement rates at one center actually went up rather than down, investigators reported at the annual meeting of the American Association for the Surgery of Trauma.

Among 42 donors managed for 20 hours or less, only six lungs (7% of possible total) and five hearts (12%) could be procured. In contrast, among 58 donors managed for more than 20 hours, 40 lungs (34%) and 26 hearts (45%) could be procured.

The surprising findings, from a prospective analysis of 100 consecutive organ donors, occurred despite the fact that there were no significant differences in attainment of preprocurement donor-management goals, said Dr. A. Britton Christmas of the department of surgery at the Carolina Medical Center in Charlotte, N.C.

A total of 133 organs, mean 3.2/donor, were obtained from donors managed 20 hours or less, and 243, mean 4.2/donor, from those managed for more than 20 hours (P less than .01). There were a mean of 2.6 organs transplanted for each donor managed 20 hours or less, and 3.7 for each managed for more than 20 (P less than .01).

“The general consensus is that early procurement removes the transplant organ from a hostile environment, but we believe that this study provides evidence to the contrary. Perhaps the reward really is worth the wait,” Dr. Christmas said.

Their original hypothesis was that shorter management times would yield higher organ procurement and transplant rates in general and heart and lung transplants in particular.

The investigators collaborated with the organ procurement organization LifeShare of the Carolinas, which covers 40 hospitals in a 20-county region. The analysis consisted of data on 100 consecutive donors from 2007 through 2008 that included age, ethnicity, blood type, mechanisms of death, management teams, organs procured, organs transplanted, donor type, donor management, reasons for non-use of organs, and donor management goals.

Causes of death were traumatic brain injuries in 44 donors, cerebrovascular accidents or strokes in 38, anoxic brain injury in 13, and other in 5.

The management goals included mean arterial pressure (60-100 mm Hg), central venous pressure (4-10 mm Hg), pH (7.30-7.45), PaO2 (greater than 100 mm Hg), serum sodium (155 mEq/L or less), serum glucose (less than 150 mg/dL), and urine output (0.5-0.3 mL/kg per hour).

Although heart and lung procurement rates were significantly higher for donors managed for longer times, there were no significant differences between the less than 20 hours vs. more than 20 hours in terms of the percentage of kidneys (85% and 91%, respectively) or livers (90% in each group) successfully harvested.

There were also no significant differences between the time groups in any of the preprocurement management goals.

The study was limited by the use of data from a single organ procurement organization, small sample size, and lack of data on the exact time of brain death, Dr. Christmas acknowledged.

The investigators speculated that the longer interval between brain death and organ procurement might permit the improvement of in situ graft function and might promote the procurement of some organs initially deemed unsuitable.

Dr. Carrie Sims, the invited discussant, said that additional data might help to explain why waiting longer than 20 hours was associated with better outcomes.

It would be helpful to know what percentage of donor organs originally deemed unacceptable became acceptable with waiting and whether the waiting period allowed for additional studies such as echocardiograms, cardiac angiograms, or bronchoscopies might have demonstrated suitability of organs, said Dr. Sims from the Trauma Center at Penn in Philadelphia. She also noted that differences in fluid balance between the two groups might explain the better rates with longer management times, since other studies have shown that overhydration of donors compromises lung procurement salvage rates.

The study was internally funded. Dr. Christmas and Dr. Sims disclosed no conflicts of interest.

BOSTON – It seems counterintuitive, but when management times of potential heart and lung donors stretched beyond 20 hours, successful organ procurement rates at one center actually went up rather than down, investigators reported at the annual meeting of the American Association for the Surgery of Trauma.

Among 42 donors managed for 20 hours or less, only six lungs (7% of possible total) and five hearts (12%) could be procured. In contrast, among 58 donors managed for more than 20 hours, 40 lungs (34%) and 26 hearts (45%) could be procured.

The surprising findings, from a prospective analysis of 100 consecutive organ donors, occurred despite the fact that there were no significant differences in attainment of preprocurement donor-management goals, said Dr. A. Britton Christmas of the department of surgery at the Carolina Medical Center in Charlotte, N.C.

A total of 133 organs, mean 3.2/donor, were obtained from donors managed 20 hours or less, and 243, mean 4.2/donor, from those managed for more than 20 hours (P less than .01). There were a mean of 2.6 organs transplanted for each donor managed 20 hours or less, and 3.7 for each managed for more than 20 (P less than .01).

“The general consensus is that early procurement removes the transplant organ from a hostile environment, but we believe that this study provides evidence to the contrary. Perhaps the reward really is worth the wait,” Dr. Christmas said.

Their original hypothesis was that shorter management times would yield higher organ procurement and transplant rates in general and heart and lung transplants in particular.

The investigators collaborated with the organ procurement organization LifeShare of the Carolinas, which covers 40 hospitals in a 20-county region. The analysis consisted of data on 100 consecutive donors from 2007 through 2008 that included age, ethnicity, blood type, mechanisms of death, management teams, organs procured, organs transplanted, donor type, donor management, reasons for non-use of organs, and donor management goals.

Causes of death were traumatic brain injuries in 44 donors, cerebrovascular accidents or strokes in 38, anoxic brain injury in 13, and other in 5.

The management goals included mean arterial pressure (60-100 mm Hg), central venous pressure (4-10 mm Hg), pH (7.30-7.45), PaO2 (greater than 100 mm Hg), serum sodium (155 mEq/L or less), serum glucose (less than 150 mg/dL), and urine output (0.5-0.3 mL/kg per hour).

Although heart and lung procurement rates were significantly higher for donors managed for longer times, there were no significant differences between the less than 20 hours vs. more than 20 hours in terms of the percentage of kidneys (85% and 91%, respectively) or livers (90% in each group) successfully harvested.

There were also no significant differences between the time groups in any of the preprocurement management goals.

The study was limited by the use of data from a single organ procurement organization, small sample size, and lack of data on the exact time of brain death, Dr. Christmas acknowledged.

The investigators speculated that the longer interval between brain death and organ procurement might permit the improvement of in situ graft function and might promote the procurement of some organs initially deemed unsuitable.

Dr. Carrie Sims, the invited discussant, said that additional data might help to explain why waiting longer than 20 hours was associated with better outcomes.

It would be helpful to know what percentage of donor organs originally deemed unacceptable became acceptable with waiting and whether the waiting period allowed for additional studies such as echocardiograms, cardiac angiograms, or bronchoscopies might have demonstrated suitability of organs, said Dr. Sims from the Trauma Center at Penn in Philadelphia. She also noted that differences in fluid balance between the two groups might explain the better rates with longer management times, since other studies have shown that overhydration of donors compromises lung procurement salvage rates.

The study was internally funded. Dr. Christmas and Dr. Sims disclosed no conflicts of interest.

BOSTON – It seems counterintuitive, but when management times of potential heart and lung donors stretched beyond 20 hours, successful organ procurement rates at one center actually went up rather than down, investigators reported at the annual meeting of the American Association for the Surgery of Trauma.

Among 42 donors managed for 20 hours or less, only six lungs (7% of possible total) and five hearts (12%) could be procured. In contrast, among 58 donors managed for more than 20 hours, 40 lungs (34%) and 26 hearts (45%) could be procured.

The surprising findings, from a prospective analysis of 100 consecutive organ donors, occurred despite the fact that there were no significant differences in attainment of preprocurement donor-management goals, said Dr. A. Britton Christmas of the department of surgery at the Carolina Medical Center in Charlotte, N.C.

A total of 133 organs, mean 3.2/donor, were obtained from donors managed 20 hours or less, and 243, mean 4.2/donor, from those managed for more than 20 hours (P less than .01). There were a mean of 2.6 organs transplanted for each donor managed 20 hours or less, and 3.7 for each managed for more than 20 (P less than .01).

“The general consensus is that early procurement removes the transplant organ from a hostile environment, but we believe that this study provides evidence to the contrary. Perhaps the reward really is worth the wait,” Dr. Christmas said.

Their original hypothesis was that shorter management times would yield higher organ procurement and transplant rates in general and heart and lung transplants in particular.

The investigators collaborated with the organ procurement organization LifeShare of the Carolinas, which covers 40 hospitals in a 20-county region. The analysis consisted of data on 100 consecutive donors from 2007 through 2008 that included age, ethnicity, blood type, mechanisms of death, management teams, organs procured, organs transplanted, donor type, donor management, reasons for non-use of organs, and donor management goals.

Causes of death were traumatic brain injuries in 44 donors, cerebrovascular accidents or strokes in 38, anoxic brain injury in 13, and other in 5.

The management goals included mean arterial pressure (60-100 mm Hg), central venous pressure (4-10 mm Hg), pH (7.30-7.45), PaO2 (greater than 100 mm Hg), serum sodium (155 mEq/L or less), serum glucose (less than 150 mg/dL), and urine output (0.5-0.3 mL/kg per hour).

Although heart and lung procurement rates were significantly higher for donors managed for longer times, there were no significant differences between the less than 20 hours vs. more than 20 hours in terms of the percentage of kidneys (85% and 91%, respectively) or livers (90% in each group) successfully harvested.

There were also no significant differences between the time groups in any of the preprocurement management goals.

The study was limited by the use of data from a single organ procurement organization, small sample size, and lack of data on the exact time of brain death, Dr. Christmas acknowledged.

The investigators speculated that the longer interval between brain death and organ procurement might permit the improvement of in situ graft function and might promote the procurement of some organs initially deemed unsuitable.

Dr. Carrie Sims, the invited discussant, said that additional data might help to explain why waiting longer than 20 hours was associated with better outcomes.

It would be helpful to know what percentage of donor organs originally deemed unacceptable became acceptable with waiting and whether the waiting period allowed for additional studies such as echocardiograms, cardiac angiograms, or bronchoscopies might have demonstrated suitability of organs, said Dr. Sims from the Trauma Center at Penn in Philadelphia. She also noted that differences in fluid balance between the two groups might explain the better rates with longer management times, since other studies have shown that overhydration of donors compromises lung procurement salvage rates.

The study was internally funded. Dr. Christmas and Dr. Sims disclosed no conflicts of interest.

BOSTON – It seems counterintuitive, but when management times of potential heart and lung donors stretched beyond 20 hours, successful organ procurement rates at one center actually went up rather than down, investigators reported at the annual meeting of the American Association for the Surgery of Trauma.

Among 42 donors managed for 20 hours or less, only six lungs (7% of possible total) and five hearts (12%) could be procured. In contrast, among 58 donors managed for more than 20 hours, 40 lungs (34%) and 26 hearts (45%) could be procured.

The surprising findings, from a prospective analysis of 100 consecutive organ donors, occurred despite the fact that there were no significant differences in attainment of preprocurement donor-management goals, said Dr. A. Britton Christmas of the department of surgery at the Carolina Medical Center in Charlotte, N.C.

A total of 133 organs, mean 3.2/donor, were obtained from donors managed 20 hours or less, and 243, mean 4.2/donor, from those managed for more than 20 hours (P less than .01). There were a mean of 2.6 organs transplanted for each donor managed 20 hours or less, and 3.7 for each managed for more than 20 (P less than .01).

“The general consensus is that early procurement removes the transplant organ from a hostile environment, but we believe that this study provides evidence to the contrary. Perhaps the reward really is worth the wait,” Dr. Christmas said.

Their original hypothesis was that shorter management times would yield higher organ procurement and transplant rates in general and heart and lung transplants in particular.

The investigators collaborated with the organ procurement organization LifeShare of the Carolinas, which covers 40 hospitals in a 20-county region. The analysis consisted of data on 100 consecutive donors from 2007 through 2008 that included age, ethnicity, blood type, mechanisms of death, management teams, organs procured, organs transplanted, donor type, donor management, reasons for non-use of organs, and donor management goals.

Causes of death were traumatic brain injuries in 44 donors, cerebrovascular accidents or strokes in 38, anoxic brain injury in 13, and other in 5.

The management goals included mean arterial pressure (60-100 mm Hg), central venous pressure (4-10 mm Hg), pH (7.30-7.45), PaO2 (greater than 100 mm Hg), serum sodium (155 mEq/L or less), serum glucose (less than 150 mg/dL), and urine output (0.5-0.3 mL/kg per hour).

Although heart and lung procurement rates were significantly higher for donors managed for longer times, there were no significant differences between the less than 20 hours vs. more than 20 hours in terms of the percentage of kidneys (85% and 91%, respectively) or livers (90% in each group) successfully harvested.

There were also no significant differences between the time groups in any of the preprocurement management goals.

The study was limited by the use of data from a single organ procurement organization, small sample size, and lack of data on the exact time of brain death, Dr. Christmas acknowledged.

The investigators speculated that the longer interval between brain death and organ procurement might permit the improvement of in situ graft function and might promote the procurement of some organs initially deemed unsuitable.

Dr. Carrie Sims, the invited discussant, said that additional data might help to explain why waiting longer than 20 hours was associated with better outcomes.

It would be helpful to know what percentage of donor organs originally deemed unacceptable became acceptable with waiting and whether the waiting period allowed for additional studies such as echocardiograms, cardiac angiograms, or bronchoscopies might have demonstrated suitability of organs, said Dr. Sims from the Trauma Center at Penn in Philadelphia. She also noted that differences in fluid balance between the two groups might explain the better rates with longer management times, since other studies have shown that overhydration of donors compromises lung procurement salvage rates.

The study was internally funded. Dr. Christmas and Dr. Sims disclosed no conflicts of interest.

BOSTON – It seems counterintuitive, but when management times of potential heart and lung donors stretched beyond 20 hours, successful organ procurement rates at one center actually went up rather than down, investigators reported at the annual meeting of the American Association for the Surgery of Trauma.

Among 42 donors managed for 20 hours or less, only six lungs (7% of possible total) and five hearts (12%) could be procured. In contrast, among 58 donors managed for more than 20 hours, 40 lungs (34%) and 26 hearts (45%) could be procured.

The surprising findings, from a prospective analysis of 100 consecutive organ donors, occurred despite the fact that there were no significant differences in attainment of preprocurement donor-management goals, said Dr. A. Britton Christmas of the department of surgery at the Carolina Medical Center in Charlotte, N.C.

A total of 133 organs, mean 3.2/donor, were obtained from donors managed 20 hours or less, and 243, mean 4.2/donor, from those managed for more than 20 hours (P less than .01). There were a mean of 2.6 organs transplanted for each donor managed 20 hours or less, and 3.7 for each managed for more than 20 (P less than .01).

“The general consensus is that early procurement removes the transplant organ from a hostile environment, but we believe that this study provides evidence to the contrary. Perhaps the reward really is worth the wait,” Dr. Christmas said.

Their original hypothesis was that shorter management times would yield higher organ procurement and transplant rates in general and heart and lung transplants in particular.

The investigators collaborated with the organ procurement organization LifeShare of the Carolinas, which covers 40 hospitals in a 20-county region. The analysis consisted of data on 100 consecutive donors from 2007 through 2008 that included age, ethnicity, blood type, mechanisms of death, management teams, organs procured, organs transplanted, donor type, donor management, reasons for non-use of organs, and donor management goals.

Causes of death were traumatic brain injuries in 44 donors, cerebrovascular accidents or strokes in 38, anoxic brain injury in 13, and other in 5.

The management goals included mean arterial pressure (60-100 mm Hg), central venous pressure (4-10 mm Hg), pH (7.30-7.45), PaO2 (greater than 100 mm Hg), serum sodium (155 mEq/L or less), serum glucose (less than 150 mg/dL), and urine output (0.5-0.3 mL/kg per hour).

Although heart and lung procurement rates were significantly higher for donors managed for longer times, there were no significant differences between the less than 20 hours vs. more than 20 hours in terms of the percentage of kidneys (85% and 91%, respectively) or livers (90% in each group) successfully harvested.

There were also no significant differences between the time groups in any of the preprocurement management goals.

The study was limited by the use of data from a single organ procurement organization, small sample size, and lack of data on the exact time of brain death, Dr. Christmas acknowledged.

The investigators speculated that the longer interval between brain death and organ procurement might permit the improvement of in situ graft function and might promote the procurement of some organs initially deemed unsuitable.

Dr. Carrie Sims, the invited discussant, said that additional data might help to explain why waiting longer than 20 hours was associated with better outcomes.

It would be helpful to know what percentage of donor organs originally deemed unacceptable became acceptable with waiting and whether the waiting period allowed for additional studies such as echocardiograms, cardiac angiograms, or bronchoscopies might have demonstrated suitability of organs, said Dr. Sims from the Trauma Center at Penn in Philadelphia. She also noted that differences in fluid balance between the two groups might explain the better rates with longer management times, since other studies have shown that overhydration of donors compromises lung procurement salvage rates.

The study was internally funded. Dr. Christmas and Dr. Sims disclosed no conflicts of interest.

BOSTON – It seems counterintuitive, but when management times of potential heart and lung donors stretched beyond 20 hours, successful organ procurement rates at one center actually went up rather than down, investigators reported at the annual meeting of the American Association for the Surgery of Trauma.

Among 42 donors managed for 20 hours or less, only six lungs (7% of possible total) and five hearts (12%) could be procured. In contrast, among 58 donors managed for more than 20 hours, 40 lungs (34%) and 26 hearts (45%) could be procured.

The surprising findings, from a prospective analysis of 100 consecutive organ donors, occurred despite the fact that there were no significant differences in attainment of preprocurement donor-management goals, said Dr. A. Britton Christmas of the department of surgery at the Carolina Medical Center in Charlotte, N.C.

A total of 133 organs, mean 3.2/donor, were obtained from donors managed 20 hours or less, and 243, mean 4.2/donor, from those managed for more than 20 hours (P less than .01). There were a mean of 2.6 organs transplanted for each donor managed 20 hours or less, and 3.7 for each managed for more than 20 (P less than .01).

“The general consensus is that early procurement removes the transplant organ from a hostile environment, but we believe that this study provides evidence to the contrary. Perhaps the reward really is worth the wait,” Dr. Christmas said.

Their original hypothesis was that shorter management times would yield higher organ procurement and transplant rates in general and heart and lung transplants in particular.

The investigators collaborated with the organ procurement organization LifeShare of the Carolinas, which covers 40 hospitals in a 20-county region. The analysis consisted of data on 100 consecutive donors from 2007 through 2008 that included age, ethnicity, blood type, mechanisms of death, management teams, organs procured, organs transplanted, donor type, donor management, reasons for non-use of organs, and donor management goals.

Causes of death were traumatic brain injuries in 44 donors, cerebrovascular accidents or strokes in 38, anoxic brain injury in 13, and other in 5.

The management goals included mean arterial pressure (60-100 mm Hg), central venous pressure (4-10 mm Hg), pH (7.30-7.45), PaO2 (greater than 100 mm Hg), serum sodium (155 mEq/L or less), serum glucose (less than 150 mg/dL), and urine output (0.5-0.3 mL/kg per hour).

Although heart and lung procurement rates were significantly higher for donors managed for longer times, there were no significant differences between the less than 20 hours vs. more than 20 hours in terms of the percentage of kidneys (85% and 91%, respectively) or livers (90% in each group) successfully harvested.

There were also no significant differences between the time groups in any of the preprocurement management goals.

The study was limited by the use of data from a single organ procurement organization, small sample size, and lack of data on the exact time of brain death, Dr. Christmas acknowledged.

The investigators speculated that the longer interval between brain death and organ procurement might permit the improvement of in situ graft function and might promote the procurement of some organs initially deemed unsuitable.

Dr. Carrie Sims, the invited discussant, said that additional data might help to explain why waiting longer than 20 hours was associated with better outcomes.

It would be helpful to know what percentage of donor organs originally deemed unacceptable became acceptable with waiting and whether the waiting period allowed for additional studies such as echocardiograms, cardiac angiograms, or bronchoscopies might have demonstrated suitability of organs, said Dr. Sims from the Trauma Center at Penn in Philadelphia. She also noted that differences in fluid balance between the two groups might explain the better rates with longer management times, since other studies have shown that overhydration of donors compromises lung procurement salvage rates.

The study was internally funded. Dr. Christmas and Dr. Sims disclosed no conflicts of interest.

BOSTON – Bacteremia secondary to pneumonia caused by carbepenem-resistant Klebsiella pneumoniae carries a high mortality rate, but combination antimicrobial regimens involving polymyxins, tigecycline, or carbepenems offer a better shot at survival compared with monotherapy, according to the findings of an observational treatment study conducted at two medical centers.

Among 41 patients infected with K. pneumoniae bacteria that produce the drug-resistant Klebsiella pneumoniae Carbepenemase (KPC), 14-day mortality was 24%, and 28-day mortality was 35%, Dr. Zubair A. Qureshi reported at the annual Interscience Conference on Antimicrobial Agents and Chemotherapy.

Mortality was increased when patients received monotherapy with either colistin (4 deaths among 7 patients) or tigecycline (3 of 5 patients). However, there were no deaths within 28 days among patients treated with either colistin or tigecycline added to any carbepenem, even when the infectious organism was reported to be nonsusceptible to carbepenems, said Dr. Qureshi of the division of internal medicine at the University of Pittsburgh Medical Center.

He noted that the preliminary findings are supported by a recent review of KPC infections that documented better clinical outcomes with combination therapy compared with monotherapy (J. Antimicrob. Chemother. 2010;6:1119-25).

KPC type beta-lactamases have been shown to confer either decreased susceptibility or resistance to virtually all beta-lactam antibiotics, including the carbapenem class agents imipenem, meropenem, and ertapenem.

Investigators at the University of Pittsburgh and St. Luke’s-Roosevelt Hospital Center in New York City conducted the single-arm observational study of treatment outcomes in patients with bacteremia due to KPC-producing K. pneumoniae. Patients were screened for the presence of KPC by reduced susceptibility to ertapenem, which was confirmed with polymerase chain reaction.

The authors looked at risk factors, antimicrobial therapy, and in-hospital mortality rates. They identified 41 patients (24 women, 17 men) with KPC-producing K. pneumoniae, with a median age of 62 years (range 25-90 years). All of the cases appeared to have been acquired in either the hospital (78%) or other health care settings such as long-term care facilities. There were no identified cases of community-acquired infections.

The source of the bacteremia was vascular catheters in 29% of the cases, pneumonia in 27%, urinary tract in 15%, intra-abdominal in 4%, and superficial wounds in 4%. The source was unknown in the remaining patients.

The primary risk factor was immunocompromised status, either from a transplant, malignancy, diabetes, connective tissue disease, chronic renal failure, or HIV infection. In all, 76% of patients had recently received antimicrobial agents, and 41% were nursing-home residents.

Deaths occurred in 7 of 11 patients with pneumonia as the source of bacteremia, 3 of 12 patients with vascular catheters as the source, 1 of 6 with urinary catheter-based infections, and 3 of 12 from other or unknown sources.

When they looked at 28-day mortality in patients who received definitive therapy, they found that any combination was associated with a significantly lower rate than monotherapy (6% vs. 59%, P = .002). The analysis did not include two patients who were lost to follow-up.

The regimens consisted of various combinations of polymyxins, tigecycline, and carbapenems.

“The combination of colistin and carbepenem appears to be superior to any other antibiotic combination, but there is a need for more observation as well as randomized clinical trials to help define the optimal treatment for KPC infections,” Dr. Qureshi said.

Disclosures: Dr. Qureshi reported having no conflicts of interest. Several of his colleagues reported receiving consulting fees from AstraZeneca, Merck, Novartis, Leo Pharmaceuticals, Three Rivers Pharmaceuticals, and/or Johnson & Johnson.

BOSTON – Bacteremia secondary to pneumonia caused by carbepenem-resistant Klebsiella pneumoniae carries a high mortality rate, but combination antimicrobial regimens involving polymyxins, tigecycline, or carbepenems offer a better shot at survival compared with monotherapy, according to the findings of an observational treatment study conducted at two medical centers.

Among 41 patients infected with K. pneumoniae bacteria that produce the drug-resistant Klebsiella pneumoniae Carbepenemase (KPC), 14-day mortality was 24%, and 28-day mortality was 35%, Dr. Zubair A. Qureshi reported at the annual Interscience Conference on Antimicrobial Agents and Chemotherapy.

Mortality was increased when patients received monotherapy with either colistin (4 deaths among 7 patients) or tigecycline (3 of 5 patients). However, there were no deaths within 28 days among patients treated with either colistin or tigecycline added to any carbepenem, even when the infectious organism was reported to be nonsusceptible to carbepenems, said Dr. Qureshi of the division of internal medicine at the University of Pittsburgh Medical Center.

He noted that the preliminary findings are supported by a recent review of KPC infections that documented better clinical outcomes with combination therapy compared with monotherapy (J. Antimicrob. Chemother. 2010;6:1119-25).

KPC type beta-lactamases have been shown to confer either decreased susceptibility or resistance to virtually all beta-lactam antibiotics, including the carbapenem class agents imipenem, meropenem, and ertapenem.

Investigators at the University of Pittsburgh and St. Luke’s-Roosevelt Hospital Center in New York City conducted the single-arm observational study of treatment outcomes in patients with bacteremia due to KPC-producing K. pneumoniae. Patients were screened for the presence of KPC by reduced susceptibility to ertapenem, which was confirmed with polymerase chain reaction.

The authors looked at risk factors, antimicrobial therapy, and in-hospital mortality rates. They identified 41 patients (24 women, 17 men) with KPC-producing K. pneumoniae, with a median age of 62 years (range 25-90 years). All of the cases appeared to have been acquired in either the hospital (78%) or other health care settings such as long-term care facilities. There were no identified cases of community-acquired infections.

The source of the bacteremia was vascular catheters in 29% of the cases, pneumonia in 27%, urinary tract in 15%, intra-abdominal in 4%, and superficial wounds in 4%. The source was unknown in the remaining patients.

The primary risk factor was immunocompromised status, either from a transplant, malignancy, diabetes, connective tissue disease, chronic renal failure, or HIV infection. In all, 76% of patients had recently received antimicrobial agents, and 41% were nursing-home residents.

Deaths occurred in 7 of 11 patients with pneumonia as the source of bacteremia, 3 of 12 patients with vascular catheters as the source, 1 of 6 with urinary catheter-based infections, and 3 of 12 from other or unknown sources.

When they looked at 28-day mortality in patients who received definitive therapy, they found that any combination was associated with a significantly lower rate than monotherapy (6% vs. 59%, P = .002). The analysis did not include two patients who were lost to follow-up.

The regimens consisted of various combinations of polymyxins, tigecycline, and carbapenems.

“The combination of colistin and carbepenem appears to be superior to any other antibiotic combination, but there is a need for more observation as well as randomized clinical trials to help define the optimal treatment for KPC infections,” Dr. Qureshi said.

Disclosures: Dr. Qureshi reported having no conflicts of interest. Several of his colleagues reported receiving consulting fees from AstraZeneca, Merck, Novartis, Leo Pharmaceuticals, Three Rivers Pharmaceuticals, and/or Johnson & Johnson.

BOSTON – Bacteremia secondary to pneumonia caused by carbepenem-resistant Klebsiella pneumoniae carries a high mortality rate, but combination antimicrobial regimens involving polymyxins, tigecycline, or carbepenems offer a better shot at survival compared with monotherapy, according to the findings of an observational treatment study conducted at two medical centers.

Among 41 patients infected with K. pneumoniae bacteria that produce the drug-resistant Klebsiella pneumoniae Carbepenemase (KPC), 14-day mortality was 24%, and 28-day mortality was 35%, Dr. Zubair A. Qureshi reported at the annual Interscience Conference on Antimicrobial Agents and Chemotherapy.

Mortality was increased when patients received monotherapy with either colistin (4 deaths among 7 patients) or tigecycline (3 of 5 patients). However, there were no deaths within 28 days among patients treated with either colistin or tigecycline added to any carbepenem, even when the infectious organism was reported to be nonsusceptible to carbepenems, said Dr. Qureshi of the division of internal medicine at the University of Pittsburgh Medical Center.

He noted that the preliminary findings are supported by a recent review of KPC infections that documented better clinical outcomes with combination therapy compared with monotherapy (J. Antimicrob. Chemother. 2010;6:1119-25).

KPC type beta-lactamases have been shown to confer either decreased susceptibility or resistance to virtually all beta-lactam antibiotics, including the carbapenem class agents imipenem, meropenem, and ertapenem.

Investigators at the University of Pittsburgh and St. Luke’s-Roosevelt Hospital Center in New York City conducted the single-arm observational study of treatment outcomes in patients with bacteremia due to KPC-producing K. pneumoniae. Patients were screened for the presence of KPC by reduced susceptibility to ertapenem, which was confirmed with polymerase chain reaction.

The authors looked at risk factors, antimicrobial therapy, and in-hospital mortality rates. They identified 41 patients (24 women, 17 men) with KPC-producing K. pneumoniae, with a median age of 62 years (range 25-90 years). All of the cases appeared to have been acquired in either the hospital (78%) or other health care settings such as long-term care facilities. There were no identified cases of community-acquired infections.

The source of the bacteremia was vascular catheters in 29% of the cases, pneumonia in 27%, urinary tract in 15%, intra-abdominal in 4%, and superficial wounds in 4%. The source was unknown in the remaining patients.

The primary risk factor was immunocompromised status, either from a transplant, malignancy, diabetes, connective tissue disease, chronic renal failure, or HIV infection. In all, 76% of patients had recently received antimicrobial agents, and 41% were nursing-home residents.

Deaths occurred in 7 of 11 patients with pneumonia as the source of bacteremia, 3 of 12 patients with vascular catheters as the source, 1 of 6 with urinary catheter-based infections, and 3 of 12 from other or unknown sources.

When they looked at 28-day mortality in patients who received definitive therapy, they found that any combination was associated with a significantly lower rate than monotherapy (6% vs. 59%, P = .002). The analysis did not include two patients who were lost to follow-up.

The regimens consisted of various combinations of polymyxins, tigecycline, and carbapenems.

“The combination of colistin and carbepenem appears to be superior to any other antibiotic combination, but there is a need for more observation as well as randomized clinical trials to help define the optimal treatment for KPC infections,” Dr. Qureshi said.

Disclosures: Dr. Qureshi reported having no conflicts of interest. Several of his colleagues reported receiving consulting fees from AstraZeneca, Merck, Novartis, Leo Pharmaceuticals, Three Rivers Pharmaceuticals, and/or Johnson & Johnson.