Perioperative Medicine

ORLANDO — Independent risk predictors have been identified that help to target the 1% of patients who develop acute renal failure following general surgery.

Eleven independent risk factors predicted acute renal failure in a logistic regression model performed by Dr. Sachin Kheterpal and his colleagues at the University of Michigan, Ann Arbor.

The researchers reviewed 150,490 operations in the American College of Surgeons' National Surgical Quality Improvement Program data set performed over a 1-year period (2005–2006) and calculated hazard ratios to compare the likelihood of acute renal failure between patients with and without each risk factor.

Preoperative renal insufficiency was associated with the greatest hazard ratio, 8.5. Other risk factors were heart failure (HR 7.6), ascites (HR 6.4), myocardial infarction within 6 months (HR 5.7), high-risk surgery (HR 3.8), aged 58 years or older (HR 3.2), hypertension requiring chronic medication (HR 3.1), male sex (HR 1.9), diabetes mellitus (HR 2.6), previous cardiac procedure (HR 2.2), and emergency surgery (HR 2.7), Dr. Kheterpal said during a poster discussion session at the annual meeting of the American Society of Anesthesiologists.

Dr. Kheterpal and his associates then took the list one step further.

They developed a “robust” risk-prediction model—an index based on the number of preoperative risk factors. The association between these risk factors and ARF incidence was as follows: one to two risk factors, 0.2% ARF incidence; three risk factors, 1% incidence; four factors, 2% incidence; five factors, 3.3% incidence; and six factors, 9% incidence.

“This is really surprising when you look at how many older patients we have, who are also men, and have hypertension and diabetes—[they] have a fairly good chance of acute renal failure,” Dr. Kheterpal said. The prediction index could be the foundation for informed consent in these patients, he added.

The investigators based their conclusions on information prospectively gathered by trained nurse data collectors from 121 centers, including community and academic centers.

“The data collection system is very detailed. It includes 30-day outcomes, even if the patient leaves and dies at another hospital,” said Dr. Kheterpal of the department of anesthesiology at the university.

Excluded from the study were patients with preexisting acute renal failure or those requiring dialysis; patients undergoing any nongeneral surgery procedures, including cases performed by vascular, cardiac, urology, ophthalmology, podiatry, or obstetric services; outpatients; and general surgery patients on whom concurrent urology procedures were performed.

Data for 68,147 operations were assessed further. Of these, 712 patients (1%) experienced acute renal failure (ARF) postoperatively. ARF was defined as progressive renal insufficiency—an increase in serum creatinine of 2 mg/dL or more above baseline—or a requirement for postoperative dialysis.

This 1% incidence of acute kidney injury is very close to a previously reported incidence of 0.8% (Anesthesiology 2007;107:892–902). The use of vasopressors and diuretics was among the factors associated with ARF in this single-center study of more than 65,000 noncardiac procedures. In addition, patients who experienced ARF had increased 30-day, 60-day, and 1-year all-cause mortality.

A meeting attendee asked Dr. Kheterpal if he and other physicians at the University of Michigan are doing anything different based on the study findings. “Yes, we are very aggressive now about hydration. And we're very aggressive now about not using diuretics intraoperatively. I've gotten into fights about that,” he said.

Having a study with nearly 70,000 general surgery cases allowed Dr. Kheterpal and his associates to do an internal validation study.

He acknowledged that there are also many limitations because the study is based on a national data set. For example, intraoperative data beyond length of each procedure were “very limited,” and there was no information on the use of preoperative medications.

“Next at our institution we will be looking at alternative serum and urinary biomarkers of ARF,” Dr. Kheterpal said.

ORLANDO — Independent risk predictors have been identified that help to target the 1% of patients who develop acute renal failure following general surgery.

Eleven independent risk factors predicted acute renal failure in a logistic regression model performed by Dr. Sachin Kheterpal and his colleagues at the University of Michigan, Ann Arbor.

The researchers reviewed 150,490 operations in the American College of Surgeons' National Surgical Quality Improvement Program data set performed over a 1-year period (2005–2006) and calculated hazard ratios to compare the likelihood of acute renal failure between patients with and without each risk factor.

Preoperative renal insufficiency was associated with the greatest hazard ratio, 8.5. Other risk factors were heart failure (HR 7.6), ascites (HR 6.4), myocardial infarction within 6 months (HR 5.7), high-risk surgery (HR 3.8), aged 58 years or older (HR 3.2), hypertension requiring chronic medication (HR 3.1), male sex (HR 1.9), diabetes mellitus (HR 2.6), previous cardiac procedure (HR 2.2), and emergency surgery (HR 2.7), Dr. Kheterpal said during a poster discussion session at the annual meeting of the American Society of Anesthesiologists.

Dr. Kheterpal and his associates then took the list one step further.

They developed a “robust” risk-prediction model—an index based on the number of preoperative risk factors. The association between these risk factors and ARF incidence was as follows: one to two risk factors, 0.2% ARF incidence; three risk factors, 1% incidence; four factors, 2% incidence; five factors, 3.3% incidence; and six factors, 9% incidence.

“This is really surprising when you look at how many older patients we have, who are also men, and have hypertension and diabetes—[they] have a fairly good chance of acute renal failure,” Dr. Kheterpal said. The prediction index could be the foundation for informed consent in these patients, he added.

The investigators based their conclusions on information prospectively gathered by trained nurse data collectors from 121 centers, including community and academic centers.

“The data collection system is very detailed. It includes 30-day outcomes, even if the patient leaves and dies at another hospital,” said Dr. Kheterpal of the department of anesthesiology at the university.

Excluded from the study were patients with preexisting acute renal failure or those requiring dialysis; patients undergoing any nongeneral surgery procedures, including cases performed by vascular, cardiac, urology, ophthalmology, podiatry, or obstetric services; outpatients; and general surgery patients on whom concurrent urology procedures were performed.

Data for 68,147 operations were assessed further. Of these, 712 patients (1%) experienced acute renal failure (ARF) postoperatively. ARF was defined as progressive renal insufficiency—an increase in serum creatinine of 2 mg/dL or more above baseline—or a requirement for postoperative dialysis.

This 1% incidence of acute kidney injury is very close to a previously reported incidence of 0.8% (Anesthesiology 2007;107:892–902). The use of vasopressors and diuretics was among the factors associated with ARF in this single-center study of more than 65,000 noncardiac procedures. In addition, patients who experienced ARF had increased 30-day, 60-day, and 1-year all-cause mortality.

A meeting attendee asked Dr. Kheterpal if he and other physicians at the University of Michigan are doing anything different based on the study findings. “Yes, we are very aggressive now about hydration. And we're very aggressive now about not using diuretics intraoperatively. I've gotten into fights about that,” he said.

Having a study with nearly 70,000 general surgery cases allowed Dr. Kheterpal and his associates to do an internal validation study.

He acknowledged that there are also many limitations because the study is based on a national data set. For example, intraoperative data beyond length of each procedure were “very limited,” and there was no information on the use of preoperative medications.

“Next at our institution we will be looking at alternative serum and urinary biomarkers of ARF,” Dr. Kheterpal said.

ORLANDO — Independent risk predictors have been identified that help to target the 1% of patients who develop acute renal failure following general surgery.

Eleven independent risk factors predicted acute renal failure in a logistic regression model performed by Dr. Sachin Kheterpal and his colleagues at the University of Michigan, Ann Arbor.

The researchers reviewed 150,490 operations in the American College of Surgeons' National Surgical Quality Improvement Program data set performed over a 1-year period (2005–2006) and calculated hazard ratios to compare the likelihood of acute renal failure between patients with and without each risk factor.

Preoperative renal insufficiency was associated with the greatest hazard ratio, 8.5. Other risk factors were heart failure (HR 7.6), ascites (HR 6.4), myocardial infarction within 6 months (HR 5.7), high-risk surgery (HR 3.8), aged 58 years or older (HR 3.2), hypertension requiring chronic medication (HR 3.1), male sex (HR 1.9), diabetes mellitus (HR 2.6), previous cardiac procedure (HR 2.2), and emergency surgery (HR 2.7), Dr. Kheterpal said during a poster discussion session at the annual meeting of the American Society of Anesthesiologists.

Dr. Kheterpal and his associates then took the list one step further.

They developed a “robust” risk-prediction model—an index based on the number of preoperative risk factors. The association between these risk factors and ARF incidence was as follows: one to two risk factors, 0.2% ARF incidence; three risk factors, 1% incidence; four factors, 2% incidence; five factors, 3.3% incidence; and six factors, 9% incidence.

“This is really surprising when you look at how many older patients we have, who are also men, and have hypertension and diabetes—[they] have a fairly good chance of acute renal failure,” Dr. Kheterpal said. The prediction index could be the foundation for informed consent in these patients, he added.

The investigators based their conclusions on information prospectively gathered by trained nurse data collectors from 121 centers, including community and academic centers.

“The data collection system is very detailed. It includes 30-day outcomes, even if the patient leaves and dies at another hospital,” said Dr. Kheterpal of the department of anesthesiology at the university.

Excluded from the study were patients with preexisting acute renal failure or those requiring dialysis; patients undergoing any nongeneral surgery procedures, including cases performed by vascular, cardiac, urology, ophthalmology, podiatry, or obstetric services; outpatients; and general surgery patients on whom concurrent urology procedures were performed.

Data for 68,147 operations were assessed further. Of these, 712 patients (1%) experienced acute renal failure (ARF) postoperatively. ARF was defined as progressive renal insufficiency—an increase in serum creatinine of 2 mg/dL or more above baseline—or a requirement for postoperative dialysis.

This 1% incidence of acute kidney injury is very close to a previously reported incidence of 0.8% (Anesthesiology 2007;107:892–902). The use of vasopressors and diuretics was among the factors associated with ARF in this single-center study of more than 65,000 noncardiac procedures. In addition, patients who experienced ARF had increased 30-day, 60-day, and 1-year all-cause mortality.

A meeting attendee asked Dr. Kheterpal if he and other physicians at the University of Michigan are doing anything different based on the study findings. “Yes, we are very aggressive now about hydration. And we're very aggressive now about not using diuretics intraoperatively. I've gotten into fights about that,” he said.

Having a study with nearly 70,000 general surgery cases allowed Dr. Kheterpal and his associates to do an internal validation study.

He acknowledged that there are also many limitations because the study is based on a national data set. For example, intraoperative data beyond length of each procedure were “very limited,” and there was no information on the use of preoperative medications.

“Next at our institution we will be looking at alternative serum and urinary biomarkers of ARF,” Dr. Kheterpal said.

WASHINGTON — When administering perioperative steroids to patients with suspected tertiary adrenal insufficiency, it's best to use the smallest possible dose for the shortest possible period of time.

“We should do our best to match the stress of the surgical procedure with the dose,” said Dr. Darrell W. Harrington, chief of the division of general internal medicine at Harbor-University of California, Los Angeles Medical Center.

The potential for adrenal crises poses a challenge in patients who are subjected to surgical stress but who are not receiving adequate corticosteroid supplementation. Data are inadequate “to allow reliable prediction of which patients are at risk for tertiary adrenal insufficiency,” he said at the annual meeting of the American College of Physicians.

“What we think we know is that there is no significant suppression if patients take a low dose [of steroids] daily.” Alternately, if patients take an every-other-day dose—usually less than 10 mg prednisone equivalent—there is little hypothalamic-pituitary-adrenal (HPA) axis suppression. Abbreviated exposures, ranging from 5 days to 3 weeks, are probably not associated with significant HPA axis suppression.

However, significant suppression does occur with high doses, superpotent topical steroids, and inhaled corticosteroids. Recovery after prolonged exposure may take as little as 5 days, or as long as a year.

“Unfortunately, there are a lot of patients who fall in between,” Dr. Harrington said. The data for these patients are inconsistent. Most patients on chronic steroids or with a history of steroid use do not have an obvious diagnosis of tertiary adrenal insufficiency. “We have to rely on clinical intuition to make this diagnosis.”

A random cortisol measurement works well at identifying patients who are likely to have HPA reserve. In addition, an adrenocorticotropic hormone (ACTH) stimulation test resulting in an incremental increase of cortisol greater than 9 mcg/dL indicates adequate HPA reserve. But biochemical abnormalities revealed by such tests “do not correlate or predict safety from a clinical events point of view,” Dr. Harrington noted.

The potential for impaired wound healing and increased infection rates related to the perioperative use of exogenous steroids makes some clinicians reluctant to adequately supplement. No clear, consistent evidence shows that wound healing is substantially delayed or impaired when patients are given short courses of perioperative steroids. And data from case series suggest that limited exposure to steroids in surgery patients is associated with a very limited increase in infection risk.

Dr. Harrington reported that he had significant financial relationships with Sanofi-Aventis, Eisai Inc., GlaxoSmithKline, and Pfizer Inc.

WASHINGTON — When administering perioperative steroids to patients with suspected tertiary adrenal insufficiency, it's best to use the smallest possible dose for the shortest possible period of time.

“We should do our best to match the stress of the surgical procedure with the dose,” said Dr. Darrell W. Harrington, chief of the division of general internal medicine at Harbor-University of California, Los Angeles Medical Center.

The potential for adrenal crises poses a challenge in patients who are subjected to surgical stress but who are not receiving adequate corticosteroid supplementation. Data are inadequate “to allow reliable prediction of which patients are at risk for tertiary adrenal insufficiency,” he said at the annual meeting of the American College of Physicians.

“What we think we know is that there is no significant suppression if patients take a low dose [of steroids] daily.” Alternately, if patients take an every-other-day dose—usually less than 10 mg prednisone equivalent—there is little hypothalamic-pituitary-adrenal (HPA) axis suppression. Abbreviated exposures, ranging from 5 days to 3 weeks, are probably not associated with significant HPA axis suppression.

However, significant suppression does occur with high doses, superpotent topical steroids, and inhaled corticosteroids. Recovery after prolonged exposure may take as little as 5 days, or as long as a year.

“Unfortunately, there are a lot of patients who fall in between,” Dr. Harrington said. The data for these patients are inconsistent. Most patients on chronic steroids or with a history of steroid use do not have an obvious diagnosis of tertiary adrenal insufficiency. “We have to rely on clinical intuition to make this diagnosis.”

A random cortisol measurement works well at identifying patients who are likely to have HPA reserve. In addition, an adrenocorticotropic hormone (ACTH) stimulation test resulting in an incremental increase of cortisol greater than 9 mcg/dL indicates adequate HPA reserve. But biochemical abnormalities revealed by such tests “do not correlate or predict safety from a clinical events point of view,” Dr. Harrington noted.

The potential for impaired wound healing and increased infection rates related to the perioperative use of exogenous steroids makes some clinicians reluctant to adequately supplement. No clear, consistent evidence shows that wound healing is substantially delayed or impaired when patients are given short courses of perioperative steroids. And data from case series suggest that limited exposure to steroids in surgery patients is associated with a very limited increase in infection risk.

Dr. Harrington reported that he had significant financial relationships with Sanofi-Aventis, Eisai Inc., GlaxoSmithKline, and Pfizer Inc.

WASHINGTON — When administering perioperative steroids to patients with suspected tertiary adrenal insufficiency, it's best to use the smallest possible dose for the shortest possible period of time.

“We should do our best to match the stress of the surgical procedure with the dose,” said Dr. Darrell W. Harrington, chief of the division of general internal medicine at Harbor-University of California, Los Angeles Medical Center.

The potential for adrenal crises poses a challenge in patients who are subjected to surgical stress but who are not receiving adequate corticosteroid supplementation. Data are inadequate “to allow reliable prediction of which patients are at risk for tertiary adrenal insufficiency,” he said at the annual meeting of the American College of Physicians.

“What we think we know is that there is no significant suppression if patients take a low dose [of steroids] daily.” Alternately, if patients take an every-other-day dose—usually less than 10 mg prednisone equivalent—there is little hypothalamic-pituitary-adrenal (HPA) axis suppression. Abbreviated exposures, ranging from 5 days to 3 weeks, are probably not associated with significant HPA axis suppression.

However, significant suppression does occur with high doses, superpotent topical steroids, and inhaled corticosteroids. Recovery after prolonged exposure may take as little as 5 days, or as long as a year.

“Unfortunately, there are a lot of patients who fall in between,” Dr. Harrington said. The data for these patients are inconsistent. Most patients on chronic steroids or with a history of steroid use do not have an obvious diagnosis of tertiary adrenal insufficiency. “We have to rely on clinical intuition to make this diagnosis.”

A random cortisol measurement works well at identifying patients who are likely to have HPA reserve. In addition, an adrenocorticotropic hormone (ACTH) stimulation test resulting in an incremental increase of cortisol greater than 9 mcg/dL indicates adequate HPA reserve. But biochemical abnormalities revealed by such tests “do not correlate or predict safety from a clinical events point of view,” Dr. Harrington noted.

The potential for impaired wound healing and increased infection rates related to the perioperative use of exogenous steroids makes some clinicians reluctant to adequately supplement. No clear, consistent evidence shows that wound healing is substantially delayed or impaired when patients are given short courses of perioperative steroids. And data from case series suggest that limited exposure to steroids in surgery patients is associated with a very limited increase in infection risk.

Dr. Harrington reported that he had significant financial relationships with Sanofi-Aventis, Eisai Inc., GlaxoSmithKline, and Pfizer Inc.

SAN DIEGO — Underweight patients have poorer overall outcomes in vascular surgery, yet morbidly obese patients have increased morbidity, primarily due to wound infections, renal complications, and thromboembolic complications, results from a multicenter analysis showed.

“There are no large studies on the effect of obesity on vascular surgery,” Dr. Eleftherios S. Xenos said at the Vascular Annual Meeting. “There is some agreement that wound infection tends to be higher in obese patients, but in terms of mortality there has not been a definite answer.”

Dr. Xenos and his associates queried the Patient Safety in Surgery Study Database for a sample of major vascular procedures performed at 14 academic medical centers in the United States between 2002 and 2004. They obtained data on 7,543 vascular surgery patients that included the National Surgical Quality Improvement Program (NSQIP) clinical definitions of patient risk factors and 30-day outcomes.

Of the 7,543 patients, 4.6% were underweight (body mass index [BMI] of less than 18.5 kg/m

Vascular procedures performed included lower extremity revascularization (24.5%), aneurysm repair (17.4%), cerebrovascular procedures (17.3%), amputations (9.4%), and “other” (31.4%). Within 30 days after surgery, 1,659 patients (22%) developed complications and 295 (3.9%) died.

As expected, risk factors for hypertension and diabetes increased with increasing BMI. However, rates of smoking, stroke, and recent weight loss—defined as more than 10% of body weight—decreased as BMI increased.

The top three NSQIP predictors of vascular surgery mortality were presence of preoperative sepsis, American Society of Anesthesiologists' physical status classification, and functional dependence. “For these three risk factors, the overweight and obese I patients had the least amount of risk, with the highest risk in the underweight category,” Dr. Xenos said.

“The overweight and obese I category patients had a significantly higher albumin level, as compared with normal class patients,” Dr. Xenos reported.

The distribution of major complications among patients was U-shaped, with a higher incidence of complications among underweight patients and the lowest among the normal weight, overweight, and obese I category patients.

Obese II and obese III category patients had a significantly higher incidence of wound infections and renal and urinary tract infection complications, while morbidly obese patients had a significantly higher incidence of thromboembolic complications.

After the researchers adjusted for age, gender, and type of operation, the 30-day mortality risk was lowest among the obese I patients (OR 0.53) and highest among the underweight patients (OR 1.48).

“Mild obesity may have an independent protective effect on nutrition, metabolic status, and improved cardiac performance,” said Dr. Xenos, who had no conflicts to disclose.

'Mild obesity may have an independent protective effect.' DR. XENOS

SAN DIEGO — Underweight patients have poorer overall outcomes in vascular surgery, yet morbidly obese patients have increased morbidity, primarily due to wound infections, renal complications, and thromboembolic complications, results from a multicenter analysis showed.

“There are no large studies on the effect of obesity on vascular surgery,” Dr. Eleftherios S. Xenos said at the Vascular Annual Meeting. “There is some agreement that wound infection tends to be higher in obese patients, but in terms of mortality there has not been a definite answer.”

Dr. Xenos and his associates queried the Patient Safety in Surgery Study Database for a sample of major vascular procedures performed at 14 academic medical centers in the United States between 2002 and 2004. They obtained data on 7,543 vascular surgery patients that included the National Surgical Quality Improvement Program (NSQIP) clinical definitions of patient risk factors and 30-day outcomes.

Of the 7,543 patients, 4.6% were underweight (body mass index [BMI] of less than 18.5 kg/m

Vascular procedures performed included lower extremity revascularization (24.5%), aneurysm repair (17.4%), cerebrovascular procedures (17.3%), amputations (9.4%), and “other” (31.4%). Within 30 days after surgery, 1,659 patients (22%) developed complications and 295 (3.9%) died.

As expected, risk factors for hypertension and diabetes increased with increasing BMI. However, rates of smoking, stroke, and recent weight loss—defined as more than 10% of body weight—decreased as BMI increased.

The top three NSQIP predictors of vascular surgery mortality were presence of preoperative sepsis, American Society of Anesthesiologists' physical status classification, and functional dependence. “For these three risk factors, the overweight and obese I patients had the least amount of risk, with the highest risk in the underweight category,” Dr. Xenos said.

“The overweight and obese I category patients had a significantly higher albumin level, as compared with normal class patients,” Dr. Xenos reported.

The distribution of major complications among patients was U-shaped, with a higher incidence of complications among underweight patients and the lowest among the normal weight, overweight, and obese I category patients.

Obese II and obese III category patients had a significantly higher incidence of wound infections and renal and urinary tract infection complications, while morbidly obese patients had a significantly higher incidence of thromboembolic complications.

After the researchers adjusted for age, gender, and type of operation, the 30-day mortality risk was lowest among the obese I patients (OR 0.53) and highest among the underweight patients (OR 1.48).

“Mild obesity may have an independent protective effect on nutrition, metabolic status, and improved cardiac performance,” said Dr. Xenos, who had no conflicts to disclose.

'Mild obesity may have an independent protective effect.' DR. XENOS

SAN DIEGO — Underweight patients have poorer overall outcomes in vascular surgery, yet morbidly obese patients have increased morbidity, primarily due to wound infections, renal complications, and thromboembolic complications, results from a multicenter analysis showed.

“There are no large studies on the effect of obesity on vascular surgery,” Dr. Eleftherios S. Xenos said at the Vascular Annual Meeting. “There is some agreement that wound infection tends to be higher in obese patients, but in terms of mortality there has not been a definite answer.”

Dr. Xenos and his associates queried the Patient Safety in Surgery Study Database for a sample of major vascular procedures performed at 14 academic medical centers in the United States between 2002 and 2004. They obtained data on 7,543 vascular surgery patients that included the National Surgical Quality Improvement Program (NSQIP) clinical definitions of patient risk factors and 30-day outcomes.

Of the 7,543 patients, 4.6% were underweight (body mass index [BMI] of less than 18.5 kg/m

Vascular procedures performed included lower extremity revascularization (24.5%), aneurysm repair (17.4%), cerebrovascular procedures (17.3%), amputations (9.4%), and “other” (31.4%). Within 30 days after surgery, 1,659 patients (22%) developed complications and 295 (3.9%) died.

As expected, risk factors for hypertension and diabetes increased with increasing BMI. However, rates of smoking, stroke, and recent weight loss—defined as more than 10% of body weight—decreased as BMI increased.

The top three NSQIP predictors of vascular surgery mortality were presence of preoperative sepsis, American Society of Anesthesiologists' physical status classification, and functional dependence. “For these three risk factors, the overweight and obese I patients had the least amount of risk, with the highest risk in the underweight category,” Dr. Xenos said.

“The overweight and obese I category patients had a significantly higher albumin level, as compared with normal class patients,” Dr. Xenos reported.

The distribution of major complications among patients was U-shaped, with a higher incidence of complications among underweight patients and the lowest among the normal weight, overweight, and obese I category patients.

Obese II and obese III category patients had a significantly higher incidence of wound infections and renal and urinary tract infection complications, while morbidly obese patients had a significantly higher incidence of thromboembolic complications.

After the researchers adjusted for age, gender, and type of operation, the 30-day mortality risk was lowest among the obese I patients (OR 0.53) and highest among the underweight patients (OR 1.48).

“Mild obesity may have an independent protective effect on nutrition, metabolic status, and improved cardiac performance,” said Dr. Xenos, who had no conflicts to disclose.

'Mild obesity may have an independent protective effect.' DR. XENOS

Tight glucose control does not benefit critically ill patients and may even be harmful because it markedly raises the risk of hypoglycemia, according to a meta-analysis.

Given these findings, “it seems appropriate that the guidelines recommending tight glucose control in all critically ill patients should be reevaluated until the results of larger, more definitive clinical trials are available,” said Dr. Renda Soylemez Wiener of the Veterans Affairs Medical Center, White River Junction, Vt., and her associates.

In 2001, a single clinical trial showed that tight glucose control reduced in-hospital mortality by a third among critically ill surgical patients. “The results of this trial were enthusiastically received and rapidly incorporated into guidelines,” and tight glucose control is now recommended for all critically ill adults and endorsed by numerous professional societies worldwide.

However, “subsequent large randomized controlled trials of tight glucose control in medical and mixed medical-surgical ICU settings … have failed to replicate this mortality benefit,” and many have reported rates of hypoglycemia as high as 40%. “Hypoglycemia is not benign in critically ill patients; it has been linked to serious neurologic events ranging from seizures to coma,” the investigators noted.

They conducted a meta-analysis of 29 controlled trials involving 8,432 adult ICU patients with a wide range of disorders who were randomly assigned to either usual care or tight glucose control—a glucose goal of less than 150 mg/dL using an insulin infusion during part or all of the ICU stay.

In-hospital mortality was not significantly different for patients kept on tight glucose control (21.6%) than for those given usual care (23.3%). There also was no difference between the two groups in the need for dialysis, Dr. Soylemez Wiener and her associates said (JAMA 2008;300:933-44).

In subgroup analyses, tight glucose control did not benefit surgical patients vs. medical patients, nor did “very tight” glucose control prove more beneficial than “moderately tight” glucose control. Sensitivity analyses of the data based on several variables that might be clinically relevant, such as disease severity, did not alter these results.

Tight glucose control did reduce the incidence of septicemia, but only among surgical ICU patients.

“On the other hand, we found clear evidence of the main harm of tight glucose control: Hypoglycemia increased roughly fivefold regardless of the ICU setting and was more common with patients receiving very tight [rather] than moderately tight glucose control,” the researchers added.

“Our meta-analysis shows that subsequent trials have not borne out the impressive results of tight glucose control promised by the initial trial” in 2001. In fact, the subjects in that initial trial received early glucose infusions and parenteral nutrition, both of which might artificially induce hyperglycemia and both of which are atypical in clinical practice. This “may have contributed to the outlying results seen in [that] trial,” the researchers noted.

In an accompanying editorial comment, Dr. Simon Finfer and Dr. Anthony Delaney of the Royal North Shore Hospital, Sydney, said that the results of this “well-conducted” and “timely” meta-analysis “may surprise many clinicians” (JAMA 2008;300:963-5).

Not only does the meta-analysis call the benefit of tight glucose control into question, it also highlights that there is no agreed standard for glycemic control, that blood glucose levels are extremely variable during the course of an ICU stay, and that bedside measurement of blood glucose can be very inaccurate, Dr. Finfer and Dr. Delaney said.

Tight glucose control does not benefit critically ill patients and may even be harmful because it markedly raises the risk of hypoglycemia, according to a meta-analysis.

Given these findings, “it seems appropriate that the guidelines recommending tight glucose control in all critically ill patients should be reevaluated until the results of larger, more definitive clinical trials are available,” said Dr. Renda Soylemez Wiener of the Veterans Affairs Medical Center, White River Junction, Vt., and her associates.

In 2001, a single clinical trial showed that tight glucose control reduced in-hospital mortality by a third among critically ill surgical patients. “The results of this trial were enthusiastically received and rapidly incorporated into guidelines,” and tight glucose control is now recommended for all critically ill adults and endorsed by numerous professional societies worldwide.

However, “subsequent large randomized controlled trials of tight glucose control in medical and mixed medical-surgical ICU settings … have failed to replicate this mortality benefit,” and many have reported rates of hypoglycemia as high as 40%. “Hypoglycemia is not benign in critically ill patients; it has been linked to serious neurologic events ranging from seizures to coma,” the investigators noted.

They conducted a meta-analysis of 29 controlled trials involving 8,432 adult ICU patients with a wide range of disorders who were randomly assigned to either usual care or tight glucose control—a glucose goal of less than 150 mg/dL using an insulin infusion during part or all of the ICU stay.

In-hospital mortality was not significantly different for patients kept on tight glucose control (21.6%) than for those given usual care (23.3%). There also was no difference between the two groups in the need for dialysis, Dr. Soylemez Wiener and her associates said (JAMA 2008;300:933-44).

In subgroup analyses, tight glucose control did not benefit surgical patients vs. medical patients, nor did “very tight” glucose control prove more beneficial than “moderately tight” glucose control. Sensitivity analyses of the data based on several variables that might be clinically relevant, such as disease severity, did not alter these results.

Tight glucose control did reduce the incidence of septicemia, but only among surgical ICU patients.

“On the other hand, we found clear evidence of the main harm of tight glucose control: Hypoglycemia increased roughly fivefold regardless of the ICU setting and was more common with patients receiving very tight [rather] than moderately tight glucose control,” the researchers added.

“Our meta-analysis shows that subsequent trials have not borne out the impressive results of tight glucose control promised by the initial trial” in 2001. In fact, the subjects in that initial trial received early glucose infusions and parenteral nutrition, both of which might artificially induce hyperglycemia and both of which are atypical in clinical practice. This “may have contributed to the outlying results seen in [that] trial,” the researchers noted.

In an accompanying editorial comment, Dr. Simon Finfer and Dr. Anthony Delaney of the Royal North Shore Hospital, Sydney, said that the results of this “well-conducted” and “timely” meta-analysis “may surprise many clinicians” (JAMA 2008;300:963-5).

Not only does the meta-analysis call the benefit of tight glucose control into question, it also highlights that there is no agreed standard for glycemic control, that blood glucose levels are extremely variable during the course of an ICU stay, and that bedside measurement of blood glucose can be very inaccurate, Dr. Finfer and Dr. Delaney said.

Tight glucose control does not benefit critically ill patients and may even be harmful because it markedly raises the risk of hypoglycemia, according to a meta-analysis.

Given these findings, “it seems appropriate that the guidelines recommending tight glucose control in all critically ill patients should be reevaluated until the results of larger, more definitive clinical trials are available,” said Dr. Renda Soylemez Wiener of the Veterans Affairs Medical Center, White River Junction, Vt., and her associates.

In 2001, a single clinical trial showed that tight glucose control reduced in-hospital mortality by a third among critically ill surgical patients. “The results of this trial were enthusiastically received and rapidly incorporated into guidelines,” and tight glucose control is now recommended for all critically ill adults and endorsed by numerous professional societies worldwide.

However, “subsequent large randomized controlled trials of tight glucose control in medical and mixed medical-surgical ICU settings … have failed to replicate this mortality benefit,” and many have reported rates of hypoglycemia as high as 40%. “Hypoglycemia is not benign in critically ill patients; it has been linked to serious neurologic events ranging from seizures to coma,” the investigators noted.

They conducted a meta-analysis of 29 controlled trials involving 8,432 adult ICU patients with a wide range of disorders who were randomly assigned to either usual care or tight glucose control—a glucose goal of less than 150 mg/dL using an insulin infusion during part or all of the ICU stay.

In-hospital mortality was not significantly different for patients kept on tight glucose control (21.6%) than for those given usual care (23.3%). There also was no difference between the two groups in the need for dialysis, Dr. Soylemez Wiener and her associates said (JAMA 2008;300:933-44).

In subgroup analyses, tight glucose control did not benefit surgical patients vs. medical patients, nor did “very tight” glucose control prove more beneficial than “moderately tight” glucose control. Sensitivity analyses of the data based on several variables that might be clinically relevant, such as disease severity, did not alter these results.

Tight glucose control did reduce the incidence of septicemia, but only among surgical ICU patients.

“On the other hand, we found clear evidence of the main harm of tight glucose control: Hypoglycemia increased roughly fivefold regardless of the ICU setting and was more common with patients receiving very tight [rather] than moderately tight glucose control,” the researchers added.

“Our meta-analysis shows that subsequent trials have not borne out the impressive results of tight glucose control promised by the initial trial” in 2001. In fact, the subjects in that initial trial received early glucose infusions and parenteral nutrition, both of which might artificially induce hyperglycemia and both of which are atypical in clinical practice. This “may have contributed to the outlying results seen in [that] trial,” the researchers noted.

In an accompanying editorial comment, Dr. Simon Finfer and Dr. Anthony Delaney of the Royal North Shore Hospital, Sydney, said that the results of this “well-conducted” and “timely” meta-analysis “may surprise many clinicians” (JAMA 2008;300:963-5).

Not only does the meta-analysis call the benefit of tight glucose control into question, it also highlights that there is no agreed standard for glycemic control, that blood glucose levels are extremely variable during the course of an ICU stay, and that bedside measurement of blood glucose can be very inaccurate, Dr. Finfer and Dr. Delaney said.

SAN FRANCISCO — Giving intraoperative insulin to cardiac surgery patients was associated with a small but significant reduction in the risk of postoperative new-onset atrial fibrillation, Dr. Michel R. Rheault said at the annual meeting of the American Society of Anesthesiologists.

The retrospective study of 9,399 patients analyzed data on all adults undergoing coronary bypass grafting or valvular procedures between January 2002 and September 2006 at the Cleveland Clinic. Of those, 5,924 had received intraoperative insulin and the rest did not. Postoperative atrial fibrillation developed in 37% of the insulin group and in 38% of the control group, reported Dr. Rheault and his associates at the clinic.

After adjusting for 44 confounding variables related to comorbid conditions, preoperative medications, operative procedures, perioperative factors, and laboratory variables, intraoperative insulin was associated with a 14% reduced risk for atrial fibrillation.

New-onset atrial fibrillation is common after cardiac surgery. Recent evidence suggesting that the inflammatory response to cardiac surgery contributes to the development of postoperative atrial fibrillation has led some surgeons to give insulin intraoperatively for its anti-inflammatory and cardioprotective effects, said Dr. Rheault.

Other changes in risk patterns were also found in the retrospective study. Patients who previously had cardiac surgery were 28% less likely to develop postoperative atrial fibrillation, whereas those undergoing a left internal thoracic artery graft had a 21% reduced risk for atrial fibrillation after surgery, compared with those undergoing other cardiac procedures.

The risk increased by 149% in patients with a history of atrial fibrillation, by 66% with every extra 10 years of age, by 28% in patients with stroke, and by 23% in patients with either cardiogenic shock or chronic obstructive pulmonary disease. The risk of postoperative atrial fibrillation was 22% higher in patients who took ?-blockers preoperatively, 19% higher in males than in females, 17% higher in patients with a history of heart failure, and 12% higher with every 5% increase in the baseline hematocrit.

The risk for postoperative atrial fibrillation was found to be 113% higher in those undergoing mitral valve replacement, 32% higher in tricuspid valvuloplasty patients, and 17% higher in those having aortic valve replacement, compared with other cardiac procedures.

The risk of postoperative atrial fibrillation increased by 27% with every unit of red blood cells transfused and by 3% with every 10-minute increase in total cross-clamp time.

SAN FRANCISCO — Giving intraoperative insulin to cardiac surgery patients was associated with a small but significant reduction in the risk of postoperative new-onset atrial fibrillation, Dr. Michel R. Rheault said at the annual meeting of the American Society of Anesthesiologists.

The retrospective study of 9,399 patients analyzed data on all adults undergoing coronary bypass grafting or valvular procedures between January 2002 and September 2006 at the Cleveland Clinic. Of those, 5,924 had received intraoperative insulin and the rest did not. Postoperative atrial fibrillation developed in 37% of the insulin group and in 38% of the control group, reported Dr. Rheault and his associates at the clinic.

After adjusting for 44 confounding variables related to comorbid conditions, preoperative medications, operative procedures, perioperative factors, and laboratory variables, intraoperative insulin was associated with a 14% reduced risk for atrial fibrillation.

New-onset atrial fibrillation is common after cardiac surgery. Recent evidence suggesting that the inflammatory response to cardiac surgery contributes to the development of postoperative atrial fibrillation has led some surgeons to give insulin intraoperatively for its anti-inflammatory and cardioprotective effects, said Dr. Rheault.

Other changes in risk patterns were also found in the retrospective study. Patients who previously had cardiac surgery were 28% less likely to develop postoperative atrial fibrillation, whereas those undergoing a left internal thoracic artery graft had a 21% reduced risk for atrial fibrillation after surgery, compared with those undergoing other cardiac procedures.

The risk increased by 149% in patients with a history of atrial fibrillation, by 66% with every extra 10 years of age, by 28% in patients with stroke, and by 23% in patients with either cardiogenic shock or chronic obstructive pulmonary disease. The risk of postoperative atrial fibrillation was 22% higher in patients who took ?-blockers preoperatively, 19% higher in males than in females, 17% higher in patients with a history of heart failure, and 12% higher with every 5% increase in the baseline hematocrit.

The risk for postoperative atrial fibrillation was found to be 113% higher in those undergoing mitral valve replacement, 32% higher in tricuspid valvuloplasty patients, and 17% higher in those having aortic valve replacement, compared with other cardiac procedures.

The risk of postoperative atrial fibrillation increased by 27% with every unit of red blood cells transfused and by 3% with every 10-minute increase in total cross-clamp time.

SAN FRANCISCO — Giving intraoperative insulin to cardiac surgery patients was associated with a small but significant reduction in the risk of postoperative new-onset atrial fibrillation, Dr. Michel R. Rheault said at the annual meeting of the American Society of Anesthesiologists.

The retrospective study of 9,399 patients analyzed data on all adults undergoing coronary bypass grafting or valvular procedures between January 2002 and September 2006 at the Cleveland Clinic. Of those, 5,924 had received intraoperative insulin and the rest did not. Postoperative atrial fibrillation developed in 37% of the insulin group and in 38% of the control group, reported Dr. Rheault and his associates at the clinic.

After adjusting for 44 confounding variables related to comorbid conditions, preoperative medications, operative procedures, perioperative factors, and laboratory variables, intraoperative insulin was associated with a 14% reduced risk for atrial fibrillation.

New-onset atrial fibrillation is common after cardiac surgery. Recent evidence suggesting that the inflammatory response to cardiac surgery contributes to the development of postoperative atrial fibrillation has led some surgeons to give insulin intraoperatively for its anti-inflammatory and cardioprotective effects, said Dr. Rheault.

Other changes in risk patterns were also found in the retrospective study. Patients who previously had cardiac surgery were 28% less likely to develop postoperative atrial fibrillation, whereas those undergoing a left internal thoracic artery graft had a 21% reduced risk for atrial fibrillation after surgery, compared with those undergoing other cardiac procedures.

The risk increased by 149% in patients with a history of atrial fibrillation, by 66% with every extra 10 years of age, by 28% in patients with stroke, and by 23% in patients with either cardiogenic shock or chronic obstructive pulmonary disease. The risk of postoperative atrial fibrillation was 22% higher in patients who took ?-blockers preoperatively, 19% higher in males than in females, 17% higher in patients with a history of heart failure, and 12% higher with every 5% increase in the baseline hematocrit.

The risk for postoperative atrial fibrillation was found to be 113% higher in those undergoing mitral valve replacement, 32% higher in tricuspid valvuloplasty patients, and 17% higher in those having aortic valve replacement, compared with other cardiac procedures.

The risk of postoperative atrial fibrillation increased by 27% with every unit of red blood cells transfused and by 3% with every 10-minute increase in total cross-clamp time.

SAN FRANCISCO — Patients with type 1 diabetes were more likely than those with type 2 diabetes to die or develop perioperative complications after undergoing total knee or hip arthroplasty, a review of 65,769 cases found.

A bivariate analysis that directly compared complications in the two diabetes groups without adjusting for confounders showed that 0.7% with type 1 diabetes and 0.3% with type 2 diabetes died in association with the total joint replacement surgery, Dr. Michael P. Bolognesi and his associates reported in a poster presentation at the annual meeting of the American Academy of Orthopaedic Surgeons.

The risk of death related to total joint arthroplasty in type 2 diabetes patients was 56% lower than that in type 1 diabetes patients.

The investigators used federal data from the 2003 National Inpatient Sample to compare the rates of complications between 8,728 patients who had type 1 diabetes and 57,041 patients who had type 2 diabetes and underwent primary and revision arthroplasties of the hip or knee from 1998 to 2003.

Their analysis was based on regression modeling to control for the potential confounding effects of age, race, gender, and median household income by zip code.

Type 2 diabetes patients were about 30% less likely than type 1 diabetes patients to develop a urinary tract infection, pneumonia, or postoperative hemorrhage; they were about 50% less likely to develop an infection related to the surgery, said Dr. Bolognesi of Duke University, Durham, N.C.

Each of the differences they found between groups was statistically significant.

Perioperative urinary infections occurred in 5% of patients with type 1 diabetes and in 3% with type 2 diabetes. Pneumonia developed in 0.8% of the type 1 diabetes group and in 0.5% of the type 2 diabetes group.

Postoperative hemorrhage after total knee or hip arthroplasty was seen in 2% of the type 1 diabetes group and in 1% of the type 2 diabetes group. Infection occurred in 0.8% of patients with type 1 diabetes and in 0.4% with type 2 diabetes.

The mean length of stay in the hospital after total joint arthroplasty was 6 days in the type 1 diabetes group at a mean inflation-adjusted cost of $33,000, and 5 days in the type 2 diabetes group at a mean inflation-adjusted cost of $31,000.

Each of these differences between groups in the bivariate analysis was statistically significant.

There were no significant differences between groups in the rates of cerebrovascular accident, mental status changes, ileus, deep vein thrombosis, pulmonary embolism, thrombophlebitis, transfusion, other wound complications, or routine discharge.

“Total joint replacement is a highly effective treatment for degenerative hip and knee conditions, despite the presence of patient comorbidities,” Dr. Bolognesi said.

Diabetes increases the risk of perioperative complications, and the different underlying pathologies of the two types of diabetes may affect the perioperative risks, he added. “As diabetes and the need for hip and knee replacements both become more prevalent, an increasingly multidisciplinary approach to the care of these patients may be one answer for improving their clinical outcomes and health care resource utilization.”

Dr. Bolognesi reported that he is a consultant to four companies that market orthopedic products, instruments, or implants: ORTHOsoft Inc., DePuy Orthopaedics Inc., Zimmer Inc., and AMEDICA Corp. He owns stock or has stock options in two of those companies, and has received research funding and nonincome support, such as equipment, from three of them.

SAN FRANCISCO — Patients with type 1 diabetes were more likely than those with type 2 diabetes to die or develop perioperative complications after undergoing total knee or hip arthroplasty, a review of 65,769 cases found.

A bivariate analysis that directly compared complications in the two diabetes groups without adjusting for confounders showed that 0.7% with type 1 diabetes and 0.3% with type 2 diabetes died in association with the total joint replacement surgery, Dr. Michael P. Bolognesi and his associates reported in a poster presentation at the annual meeting of the American Academy of Orthopaedic Surgeons.

The risk of death related to total joint arthroplasty in type 2 diabetes patients was 56% lower than that in type 1 diabetes patients.

The investigators used federal data from the 2003 National Inpatient Sample to compare the rates of complications between 8,728 patients who had type 1 diabetes and 57,041 patients who had type 2 diabetes and underwent primary and revision arthroplasties of the hip or knee from 1998 to 2003.

Their analysis was based on regression modeling to control for the potential confounding effects of age, race, gender, and median household income by zip code.

Type 2 diabetes patients were about 30% less likely than type 1 diabetes patients to develop a urinary tract infection, pneumonia, or postoperative hemorrhage; they were about 50% less likely to develop an infection related to the surgery, said Dr. Bolognesi of Duke University, Durham, N.C.

Each of the differences they found between groups was statistically significant.

Perioperative urinary infections occurred in 5% of patients with type 1 diabetes and in 3% with type 2 diabetes. Pneumonia developed in 0.8% of the type 1 diabetes group and in 0.5% of the type 2 diabetes group.

Postoperative hemorrhage after total knee or hip arthroplasty was seen in 2% of the type 1 diabetes group and in 1% of the type 2 diabetes group. Infection occurred in 0.8% of patients with type 1 diabetes and in 0.4% with type 2 diabetes.

The mean length of stay in the hospital after total joint arthroplasty was 6 days in the type 1 diabetes group at a mean inflation-adjusted cost of $33,000, and 5 days in the type 2 diabetes group at a mean inflation-adjusted cost of $31,000.

Each of these differences between groups in the bivariate analysis was statistically significant.

There were no significant differences between groups in the rates of cerebrovascular accident, mental status changes, ileus, deep vein thrombosis, pulmonary embolism, thrombophlebitis, transfusion, other wound complications, or routine discharge.

“Total joint replacement is a highly effective treatment for degenerative hip and knee conditions, despite the presence of patient comorbidities,” Dr. Bolognesi said.

Diabetes increases the risk of perioperative complications, and the different underlying pathologies of the two types of diabetes may affect the perioperative risks, he added. “As diabetes and the need for hip and knee replacements both become more prevalent, an increasingly multidisciplinary approach to the care of these patients may be one answer for improving their clinical outcomes and health care resource utilization.”

Dr. Bolognesi reported that he is a consultant to four companies that market orthopedic products, instruments, or implants: ORTHOsoft Inc., DePuy Orthopaedics Inc., Zimmer Inc., and AMEDICA Corp. He owns stock or has stock options in two of those companies, and has received research funding and nonincome support, such as equipment, from three of them.

SAN FRANCISCO — Patients with type 1 diabetes were more likely than those with type 2 diabetes to die or develop perioperative complications after undergoing total knee or hip arthroplasty, a review of 65,769 cases found.

A bivariate analysis that directly compared complications in the two diabetes groups without adjusting for confounders showed that 0.7% with type 1 diabetes and 0.3% with type 2 diabetes died in association with the total joint replacement surgery, Dr. Michael P. Bolognesi and his associates reported in a poster presentation at the annual meeting of the American Academy of Orthopaedic Surgeons.

The risk of death related to total joint arthroplasty in type 2 diabetes patients was 56% lower than that in type 1 diabetes patients.

The investigators used federal data from the 2003 National Inpatient Sample to compare the rates of complications between 8,728 patients who had type 1 diabetes and 57,041 patients who had type 2 diabetes and underwent primary and revision arthroplasties of the hip or knee from 1998 to 2003.

Their analysis was based on regression modeling to control for the potential confounding effects of age, race, gender, and median household income by zip code.

Type 2 diabetes patients were about 30% less likely than type 1 diabetes patients to develop a urinary tract infection, pneumonia, or postoperative hemorrhage; they were about 50% less likely to develop an infection related to the surgery, said Dr. Bolognesi of Duke University, Durham, N.C.

Each of the differences they found between groups was statistically significant.

Perioperative urinary infections occurred in 5% of patients with type 1 diabetes and in 3% with type 2 diabetes. Pneumonia developed in 0.8% of the type 1 diabetes group and in 0.5% of the type 2 diabetes group.

Postoperative hemorrhage after total knee or hip arthroplasty was seen in 2% of the type 1 diabetes group and in 1% of the type 2 diabetes group. Infection occurred in 0.8% of patients with type 1 diabetes and in 0.4% with type 2 diabetes.

The mean length of stay in the hospital after total joint arthroplasty was 6 days in the type 1 diabetes group at a mean inflation-adjusted cost of $33,000, and 5 days in the type 2 diabetes group at a mean inflation-adjusted cost of $31,000.

Each of these differences between groups in the bivariate analysis was statistically significant.

There were no significant differences between groups in the rates of cerebrovascular accident, mental status changes, ileus, deep vein thrombosis, pulmonary embolism, thrombophlebitis, transfusion, other wound complications, or routine discharge.

“Total joint replacement is a highly effective treatment for degenerative hip and knee conditions, despite the presence of patient comorbidities,” Dr. Bolognesi said.

Diabetes increases the risk of perioperative complications, and the different underlying pathologies of the two types of diabetes may affect the perioperative risks, he added. “As diabetes and the need for hip and knee replacements both become more prevalent, an increasingly multidisciplinary approach to the care of these patients may be one answer for improving their clinical outcomes and health care resource utilization.”

Dr. Bolognesi reported that he is a consultant to four companies that market orthopedic products, instruments, or implants: ORTHOsoft Inc., DePuy Orthopaedics Inc., Zimmer Inc., and AMEDICA Corp. He owns stock or has stock options in two of those companies, and has received research funding and nonincome support, such as equipment, from three of them.

SAN FRANCISCO — Complications after total knee or hip arthroplasty in patients with diabetes were significantly more common in those with higher hemoglobin A_1c levels, a retrospective study of 119 patients found.

The overall rate of medical and surgical complications was more than 50% in patients with a hemoglobin A_1c (HbA_1c) level greater than 7%, and less than 40% in those with an HbA_1c level below 7%, study investigators reported.

“We believe that the HbA_1c test should be a routine preoperative test ordered for diabetes patients prior to total joint arthroplasty,” Dr. Yossef C. Blum said during a poster session at the annual meeting of the American Academy of Orthopaedic Surgeons. “Patients with significantly elevated HbA_1c levels should have their glycemic control optimized prior to undergoing total hip arthroplasty or total knee arthroplasty, as well as in the perioperative period.”

In a review of inpatient and outpatient charts of total knee or hip arthroplasty performed by a single surgeon at one institution from 2000 to 2007, Dr. Blum and his associates found 199 patients whose HbA_1c level had been measured in the year before surgery or within 3 months after the surgery. Patients were excluded from the study if they had conditions other than diabetes that led to an immunosuppressed state, such as HIV or rheumatoid arthritis.

Patients did not have to have a diagnosis of diabetes to be included in the study—just an HbA_1c measurement—because up to a third of people with diabetes do not have a formal diagnosis, he and his associates reported.

In all, 73% of the patients underwent total knee arthroplasty and 27% had total hip arthroplasty.

Patients had a mean age of 68 years. The cohort was 76% male, 34% white, 34% black, 23% Hispanic, and 9% other races/ethnicities. Their mean body mass index was 34 kg/m

The investigators performed a multivariate analysis looking for associations between HbA_1c levels and outcomes within 3 months of the surgery, said Dr. Blum of Montefiore Medical Center, New York.

Higher HbA_1c levels were significantly associated with an increased risk for any complications, surgical site complications, and wound complications after total knee or hip arthroplasty, Dr. Blum and his associates reported.

Only four surgical site infections occurred—too few to demonstrate a specific association between HbA_1c levels and wound infection—but “it is notable that three of four infections occurred in patients with an HbA_1c [level] above 7.5%,” he said.

The current study found no association between HbA_1c level and the risk of non-surgical-site infections, urinary retention, or discharge after surgery to an inpatient facility.

Overall, 43% of the patients developed medical or surgical complications.

“Future studies with increased numbers of patients may help determine a cutoff HbA_1c level above which total hip arthroplasty or total knee arthroplasty can be considered too high risk,” Dr. Blum said.

Recommendations from the American Diabetes Association set a treatment goal of an HbA_1c level below 7%.

A 2003 review by other investigators of 290 diabetes patients who underwent noncardiac surgeries found that those with an HbA_1c level above 7% had a statistically significant increased risk for postoperative complications, Dr. Blum said.

There have been few studies in the past 2 decades on the results of total knee arthroplasty in diabetes patients, and even fewer studies on the results of total hip arthroplasty in diabetes patients, Dr. Blum noted. Some reports suggest a risk of 1%–7% for deep infection, and overall wound complication rates of 1%–12%.

Diabetes patients may be more prone to perioperative urinary tract infection, pneumonia, sepsis, cardiac problems, and postoperative neuropathy.

Studies suggest diabetes patients are more likely than nondiabetics to be discharged to an inpatient facility after total knee arthroplasty.

SAN FRANCISCO — Complications after total knee or hip arthroplasty in patients with diabetes were significantly more common in those with higher hemoglobin A_1c levels, a retrospective study of 119 patients found.

The overall rate of medical and surgical complications was more than 50% in patients with a hemoglobin A_1c (HbA_1c) level greater than 7%, and less than 40% in those with an HbA_1c level below 7%, study investigators reported.

“We believe that the HbA_1c test should be a routine preoperative test ordered for diabetes patients prior to total joint arthroplasty,” Dr. Yossef C. Blum said during a poster session at the annual meeting of the American Academy of Orthopaedic Surgeons. “Patients with significantly elevated HbA_1c levels should have their glycemic control optimized prior to undergoing total hip arthroplasty or total knee arthroplasty, as well as in the perioperative period.”

In a review of inpatient and outpatient charts of total knee or hip arthroplasty performed by a single surgeon at one institution from 2000 to 2007, Dr. Blum and his associates found 199 patients whose HbA_1c level had been measured in the year before surgery or within 3 months after the surgery. Patients were excluded from the study if they had conditions other than diabetes that led to an immunosuppressed state, such as HIV or rheumatoid arthritis.

Patients did not have to have a diagnosis of diabetes to be included in the study—just an HbA_1c measurement—because up to a third of people with diabetes do not have a formal diagnosis, he and his associates reported.

In all, 73% of the patients underwent total knee arthroplasty and 27% had total hip arthroplasty.

Patients had a mean age of 68 years. The cohort was 76% male, 34% white, 34% black, 23% Hispanic, and 9% other races/ethnicities. Their mean body mass index was 34 kg/m

The investigators performed a multivariate analysis looking for associations between HbA_1c levels and outcomes within 3 months of the surgery, said Dr. Blum of Montefiore Medical Center, New York.

Higher HbA_1c levels were significantly associated with an increased risk for any complications, surgical site complications, and wound complications after total knee or hip arthroplasty, Dr. Blum and his associates reported.

Only four surgical site infections occurred—too few to demonstrate a specific association between HbA_1c levels and wound infection—but “it is notable that three of four infections occurred in patients with an HbA_1c [level] above 7.5%,” he said.

The current study found no association between HbA_1c level and the risk of non-surgical-site infections, urinary retention, or discharge after surgery to an inpatient facility.

Overall, 43% of the patients developed medical or surgical complications.

“Future studies with increased numbers of patients may help determine a cutoff HbA_1c level above which total hip arthroplasty or total knee arthroplasty can be considered too high risk,” Dr. Blum said.

Recommendations from the American Diabetes Association set a treatment goal of an HbA_1c level below 7%.

A 2003 review by other investigators of 290 diabetes patients who underwent noncardiac surgeries found that those with an HbA_1c level above 7% had a statistically significant increased risk for postoperative complications, Dr. Blum said.

There have been few studies in the past 2 decades on the results of total knee arthroplasty in diabetes patients, and even fewer studies on the results of total hip arthroplasty in diabetes patients, Dr. Blum noted. Some reports suggest a risk of 1%–7% for deep infection, and overall wound complication rates of 1%–12%.

Diabetes patients may be more prone to perioperative urinary tract infection, pneumonia, sepsis, cardiac problems, and postoperative neuropathy.

Studies suggest diabetes patients are more likely than nondiabetics to be discharged to an inpatient facility after total knee arthroplasty.

SAN FRANCISCO — Complications after total knee or hip arthroplasty in patients with diabetes were significantly more common in those with higher hemoglobin A_1c levels, a retrospective study of 119 patients found.

The overall rate of medical and surgical complications was more than 50% in patients with a hemoglobin A_1c (HbA_1c) level greater than 7%, and less than 40% in those with an HbA_1c level below 7%, study investigators reported.

“We believe that the HbA_1c test should be a routine preoperative test ordered for diabetes patients prior to total joint arthroplasty,” Dr. Yossef C. Blum said during a poster session at the annual meeting of the American Academy of Orthopaedic Surgeons. “Patients with significantly elevated HbA_1c levels should have their glycemic control optimized prior to undergoing total hip arthroplasty or total knee arthroplasty, as well as in the perioperative period.”

In a review of inpatient and outpatient charts of total knee or hip arthroplasty performed by a single surgeon at one institution from 2000 to 2007, Dr. Blum and his associates found 199 patients whose HbA_1c level had been measured in the year before surgery or within 3 months after the surgery. Patients were excluded from the study if they had conditions other than diabetes that led to an immunosuppressed state, such as HIV or rheumatoid arthritis.

Patients did not have to have a diagnosis of diabetes to be included in the study—just an HbA_1c measurement—because up to a third of people with diabetes do not have a formal diagnosis, he and his associates reported.

In all, 73% of the patients underwent total knee arthroplasty and 27% had total hip arthroplasty.

Patients had a mean age of 68 years. The cohort was 76% male, 34% white, 34% black, 23% Hispanic, and 9% other races/ethnicities. Their mean body mass index was 34 kg/m

The investigators performed a multivariate analysis looking for associations between HbA_1c levels and outcomes within 3 months of the surgery, said Dr. Blum of Montefiore Medical Center, New York.

Higher HbA_1c levels were significantly associated with an increased risk for any complications, surgical site complications, and wound complications after total knee or hip arthroplasty, Dr. Blum and his associates reported.

Only four surgical site infections occurred—too few to demonstrate a specific association between HbA_1c levels and wound infection—but “it is notable that three of four infections occurred in patients with an HbA_1c [level] above 7.5%,” he said.

The current study found no association between HbA_1c level and the risk of non-surgical-site infections, urinary retention, or discharge after surgery to an inpatient facility.

Overall, 43% of the patients developed medical or surgical complications.

“Future studies with increased numbers of patients may help determine a cutoff HbA_1c level above which total hip arthroplasty or total knee arthroplasty can be considered too high risk,” Dr. Blum said.

Recommendations from the American Diabetes Association set a treatment goal of an HbA_1c level below 7%.

A 2003 review by other investigators of 290 diabetes patients who underwent noncardiac surgeries found that those with an HbA_1c level above 7% had a statistically significant increased risk for postoperative complications, Dr. Blum said.

There have been few studies in the past 2 decades on the results of total knee arthroplasty in diabetes patients, and even fewer studies on the results of total hip arthroplasty in diabetes patients, Dr. Blum noted. Some reports suggest a risk of 1%–7% for deep infection, and overall wound complication rates of 1%–12%.

Diabetes patients may be more prone to perioperative urinary tract infection, pneumonia, sepsis, cardiac problems, and postoperative neuropathy.

Studies suggest diabetes patients are more likely than nondiabetics to be discharged to an inpatient facility after total knee arthroplasty.

SAN FRANCISCO — Instilling purified capsaicin into a surgical wound after open mesh groin hernia repair and before wound closure reduced postoperative pain scores, compared with placebo, for 3 days after surgery, according to a randomized, double-blind, placebo-controlled study of 41 men.

There was no significant difference between groups, however, in the primary end point: average daily pain scores during the first week after surgery, Dr. Eske K. Aasvang reported in a poster presentation at the annual meeting of the American Society of Anesthesiologists.

Average pain scores, as assessed on a 100-point visual analog scale, were less than 15 after postoperative day 4 in both groups—too low to show a treatment difference in the second half of the week, suggested Dr. Aasvang of the University of Copenhagen and his associates.

Dr. Aasvang received a salary from the maker of the capsaicin formulation, Anesiva Inc., which also funded the study.

Patients received a single intraoperative instillation of 15 mL of medication containing either placebo or 1,000 mcg of Adlea, a formulation consisting of more than 98% capsaicin. All patients also were given acetaminophen and ibuprofen for postoperative analgesia and were allowed to take tramadol if needed.

The morning after surgery, median pain scores on the visual analog scale were approximately 13 in the capsaicin group and 33 in the placebo group. Median pain scores in the capsaicin group and the placebo group on the second morning after surgery were approximately 15 and 25, respectively, and on the third morning were approximately 7 and 17, respectively.

Four patients in the capsaicin group and none in the placebo group developed treatment-related adverse events, including increased blood pressure (two patients), decreased heart rate (one), and abnormal skin odor (one). Dr. Aasvang described the capsaicin formulation as “generally well tolerated” in this study.

All patients except one in the placebo group had normal wound healing in the week after surgery. One patient in the placebo group had surgical wound drainage, a serious adverse event that was thought to be unrelated to the study treatment.

The study results suggest that purified capsaicin may have a role to play in reducing acute postoperative pain, the investigators concluded.

SAN FRANCISCO — Instilling purified capsaicin into a surgical wound after open mesh groin hernia repair and before wound closure reduced postoperative pain scores, compared with placebo, for 3 days after surgery, according to a randomized, double-blind, placebo-controlled study of 41 men.

There was no significant difference between groups, however, in the primary end point: average daily pain scores during the first week after surgery, Dr. Eske K. Aasvang reported in a poster presentation at the annual meeting of the American Society of Anesthesiologists.

Average pain scores, as assessed on a 100-point visual analog scale, were less than 15 after postoperative day 4 in both groups—too low to show a treatment difference in the second half of the week, suggested Dr. Aasvang of the University of Copenhagen and his associates.

Dr. Aasvang received a salary from the maker of the capsaicin formulation, Anesiva Inc., which also funded the study.

Patients received a single intraoperative instillation of 15 mL of medication containing either placebo or 1,000 mcg of Adlea, a formulation consisting of more than 98% capsaicin. All patients also were given acetaminophen and ibuprofen for postoperative analgesia and were allowed to take tramadol if needed.

The morning after surgery, median pain scores on the visual analog scale were approximately 13 in the capsaicin group and 33 in the placebo group. Median pain scores in the capsaicin group and the placebo group on the second morning after surgery were approximately 15 and 25, respectively, and on the third morning were approximately 7 and 17, respectively.

Four patients in the capsaicin group and none in the placebo group developed treatment-related adverse events, including increased blood pressure (two patients), decreased heart rate (one), and abnormal skin odor (one). Dr. Aasvang described the capsaicin formulation as “generally well tolerated” in this study.

All patients except one in the placebo group had normal wound healing in the week after surgery. One patient in the placebo group had surgical wound drainage, a serious adverse event that was thought to be unrelated to the study treatment.

The study results suggest that purified capsaicin may have a role to play in reducing acute postoperative pain, the investigators concluded.

SAN FRANCISCO — Instilling purified capsaicin into a surgical wound after open mesh groin hernia repair and before wound closure reduced postoperative pain scores, compared with placebo, for 3 days after surgery, according to a randomized, double-blind, placebo-controlled study of 41 men.

There was no significant difference between groups, however, in the primary end point: average daily pain scores during the first week after surgery, Dr. Eske K. Aasvang reported in a poster presentation at the annual meeting of the American Society of Anesthesiologists.

Average pain scores, as assessed on a 100-point visual analog scale, were less than 15 after postoperative day 4 in both groups—too low to show a treatment difference in the second half of the week, suggested Dr. Aasvang of the University of Copenhagen and his associates.

Dr. Aasvang received a salary from the maker of the capsaicin formulation, Anesiva Inc., which also funded the study.

Patients received a single intraoperative instillation of 15 mL of medication containing either placebo or 1,000 mcg of Adlea, a formulation consisting of more than 98% capsaicin. All patients also were given acetaminophen and ibuprofen for postoperative analgesia and were allowed to take tramadol if needed.

The morning after surgery, median pain scores on the visual analog scale were approximately 13 in the capsaicin group and 33 in the placebo group. Median pain scores in the capsaicin group and the placebo group on the second morning after surgery were approximately 15 and 25, respectively, and on the third morning were approximately 7 and 17, respectively.

Four patients in the capsaicin group and none in the placebo group developed treatment-related adverse events, including increased blood pressure (two patients), decreased heart rate (one), and abnormal skin odor (one). Dr. Aasvang described the capsaicin formulation as “generally well tolerated” in this study.

All patients except one in the placebo group had normal wound healing in the week after surgery. One patient in the placebo group had surgical wound drainage, a serious adverse event that was thought to be unrelated to the study treatment.

The study results suggest that purified capsaicin may have a role to play in reducing acute postoperative pain, the investigators concluded.

BOSTON — A well-established risk-scoring system is turning out to have helpful new uses for gauging the preoperative risk associated with a variety of procedures.

The Model for End-Stage Liver Disease (MELD) score was first devised to assess progression in patients with cirrhotic liver disease. But the MELD score can also be used to integrate noninvasively collected data on a patient's hepatic, renal, and coagulopathy states, making it well suited to quickly assess a patient's risk for multisystem organ dysfunction and other adverse surgical outcomes.

When applied retrospectively in 211 patients receiving a left ventricular assist device (LVAD) at one center, a modified MELD score predicted the risk for death, renal failure, and right ventricular failure following surgery, as well as patients' perioperative and postoperative need for blood products and their hospital length of stay, Dr. Jennifer C. Matthews reported at the annual meeting of the International Society for Heart and Lung Transplantation.

In other recent studies, the MELD score has been used successfully to predict the risk for adverse outcomes in patients undergoing abdominal surgery and certain cardiac procedures such as coronary bypass and valve repair or replacement, said Dr. Matthews, a cardiologist at the University of Michigan, Ann Arbor.

“I stole the score and applied it to a different population and different organ systems,” she said.

MELD scores are determined by plugging patients' serum creatinine and bilirubin levels and their international normalized ratios (INRs) into a formula that's available on the Internet. In her study, Dr. Matthews used values obtained within 24 hours before LVAD placement surgery. She used a version of the MELD score formula that has been modified by the United Network for Organ Sharing.

Her study used data collected on all 211 patients who received an LVAD at the University of Michigan during October 1996-February 2007. Their average age was 50, their average serum values were a creatinine of 1.5 mg/dL and a bilirubin of 1.8 mg/dL, and their average INR was 1.2. Their average MELD score was 13.7. Perioperative deaths occurred in 29 patients.

A multivariate analysis showed that each 1-point rise in patients' MELD scores boosted their risk of operative death by 20%. A MELD score of 17 or greater was seen in the sickest quartile of patients. Patients in this subgroup had a threefold increased risk of death, and about a fivefold increased risk for both renal failure and right ventricular failure, compared with patients whose MELD scores were less than 17.

Each 1-point rise in MELD score was linked with about a half-day longer ICU stay. Higher MELD scores were also linked with a greater need for blood products (including use of red cells, platelets, plasma, or cryoprecipitate) during or within 24 hours following surgery.

In addition to emerging as an effective prognostic tool, the MELD score can guide physicians to take preoperative measures that might improve a patient's score, such as optimizing right ventricular filling pressure, correcting coagulopathy by improved nutrition or vitamin K supplementation, and using mechanical circulatory support early on.

Dr. Matthews cautioned that it's premature to use MELD scoring to assess patients scheduled to receive an LVAD, because she has not yet shown that improving patients' scores preoperatively will yield better outcomes. She has a project underway to further validate the score's prognostic ability in another set of LVAD recipients, she said in an interview.

BOSTON — A well-established risk-scoring system is turning out to have helpful new uses for gauging the preoperative risk associated with a variety of procedures.

The Model for End-Stage Liver Disease (MELD) score was first devised to assess progression in patients with cirrhotic liver disease. But the MELD score can also be used to integrate noninvasively collected data on a patient's hepatic, renal, and coagulopathy states, making it well suited to quickly assess a patient's risk for multisystem organ dysfunction and other adverse surgical outcomes.

When applied retrospectively in 211 patients receiving a left ventricular assist device (LVAD) at one center, a modified MELD score predicted the risk for death, renal failure, and right ventricular failure following surgery, as well as patients' perioperative and postoperative need for blood products and their hospital length of stay, Dr. Jennifer C. Matthews reported at the annual meeting of the International Society for Heart and Lung Transplantation.

In other recent studies, the MELD score has been used successfully to predict the risk for adverse outcomes in patients undergoing abdominal surgery and certain cardiac procedures such as coronary bypass and valve repair or replacement, said Dr. Matthews, a cardiologist at the University of Michigan, Ann Arbor.

“I stole the score and applied it to a different population and different organ systems,” she said.

MELD scores are determined by plugging patients' serum creatinine and bilirubin levels and their international normalized ratios (INRs) into a formula that's available on the Internet. In her study, Dr. Matthews used values obtained within 24 hours before LVAD placement surgery. She used a version of the MELD score formula that has been modified by the United Network for Organ Sharing.

Her study used data collected on all 211 patients who received an LVAD at the University of Michigan during October 1996-February 2007. Their average age was 50, their average serum values were a creatinine of 1.5 mg/dL and a bilirubin of 1.8 mg/dL, and their average INR was 1.2. Their average MELD score was 13.7. Perioperative deaths occurred in 29 patients.

A multivariate analysis showed that each 1-point rise in patients' MELD scores boosted their risk of operative death by 20%. A MELD score of 17 or greater was seen in the sickest quartile of patients. Patients in this subgroup had a threefold increased risk of death, and about a fivefold increased risk for both renal failure and right ventricular failure, compared with patients whose MELD scores were less than 17.

Each 1-point rise in MELD score was linked with about a half-day longer ICU stay. Higher MELD scores were also linked with a greater need for blood products (including use of red cells, platelets, plasma, or cryoprecipitate) during or within 24 hours following surgery.

In addition to emerging as an effective prognostic tool, the MELD score can guide physicians to take preoperative measures that might improve a patient's score, such as optimizing right ventricular filling pressure, correcting coagulopathy by improved nutrition or vitamin K supplementation, and using mechanical circulatory support early on.

Dr. Matthews cautioned that it's premature to use MELD scoring to assess patients scheduled to receive an LVAD, because she has not yet shown that improving patients' scores preoperatively will yield better outcomes. She has a project underway to further validate the score's prognostic ability in another set of LVAD recipients, she said in an interview.

BOSTON — A well-established risk-scoring system is turning out to have helpful new uses for gauging the preoperative risk associated with a variety of procedures.

The Model for End-Stage Liver Disease (MELD) score was first devised to assess progression in patients with cirrhotic liver disease. But the MELD score can also be used to integrate noninvasively collected data on a patient's hepatic, renal, and coagulopathy states, making it well suited to quickly assess a patient's risk for multisystem organ dysfunction and other adverse surgical outcomes.

When applied retrospectively in 211 patients receiving a left ventricular assist device (LVAD) at one center, a modified MELD score predicted the risk for death, renal failure, and right ventricular failure following surgery, as well as patients' perioperative and postoperative need for blood products and their hospital length of stay, Dr. Jennifer C. Matthews reported at the annual meeting of the International Society for Heart and Lung Transplantation.

In other recent studies, the MELD score has been used successfully to predict the risk for adverse outcomes in patients undergoing abdominal surgery and certain cardiac procedures such as coronary bypass and valve repair or replacement, said Dr. Matthews, a cardiologist at the University of Michigan, Ann Arbor.

“I stole the score and applied it to a different population and different organ systems,” she said.

MELD scores are determined by plugging patients' serum creatinine and bilirubin levels and their international normalized ratios (INRs) into a formula that's available on the Internet. In her study, Dr. Matthews used values obtained within 24 hours before LVAD placement surgery. She used a version of the MELD score formula that has been modified by the United Network for Organ Sharing.

Her study used data collected on all 211 patients who received an LVAD at the University of Michigan during October 1996-February 2007. Their average age was 50, their average serum values were a creatinine of 1.5 mg/dL and a bilirubin of 1.8 mg/dL, and their average INR was 1.2. Their average MELD score was 13.7. Perioperative deaths occurred in 29 patients.

A multivariate analysis showed that each 1-point rise in patients' MELD scores boosted their risk of operative death by 20%. A MELD score of 17 or greater was seen in the sickest quartile of patients. Patients in this subgroup had a threefold increased risk of death, and about a fivefold increased risk for both renal failure and right ventricular failure, compared with patients whose MELD scores were less than 17.

Each 1-point rise in MELD score was linked with about a half-day longer ICU stay. Higher MELD scores were also linked with a greater need for blood products (including use of red cells, platelets, plasma, or cryoprecipitate) during or within 24 hours following surgery.

In addition to emerging as an effective prognostic tool, the MELD score can guide physicians to take preoperative measures that might improve a patient's score, such as optimizing right ventricular filling pressure, correcting coagulopathy by improved nutrition or vitamin K supplementation, and using mechanical circulatory support early on.

Dr. Matthews cautioned that it's premature to use MELD scoring to assess patients scheduled to receive an LVAD, because she has not yet shown that improving patients' scores preoperatively will yield better outcomes. She has a project underway to further validate the score's prognostic ability in another set of LVAD recipients, she said in an interview.

CHICAGO — Obesity was not a risk factor for postoperative mortality or major complications in patients undergoing major intra-abdominal cancer surgery, according to data from a prospective, multicenter risk-adjusted cohort study of 2,258 patients.

As observed in previous studies, however, obesity was a risk factor for minor complications, primarily wound infections.

Importantly, patients who were underweight had a fivefold increased risk of postoperative death, perhaps as a consequence of their underlying poor nutritional status or impaired immunity, investigator Dr. John T. Mullen reported at a symposium sponsored by the Society of Surgical Oncology.

“The prejudice that overweight and obese patients are at increased risk for serious adverse outcomes after major cancer surgery is not justified,” he said. “Paradoxically, overweight status and mild obesity status may be protective of low mortality.”

This possible “obesity paradox” warrants further study, whereas underweight patients might benefit from perioperative nutritional supplementation to mitigate their increased risk of mortality, Dr. Mullen, a surgical oncologist at Beth Israel Deaconess Medical Center, Boston, and his associates concluded.

Underweight patients in the study were significantly more likely to have lost more than 10% of their body weight in the 6 months prior to surgery and to have a lower mean preoperative serum albumin level than other patients have.

Using National Institutes of Health-defined body mass index (BMI) classes, 55 patients were stratified as underweight (BMI up to 18.5 kg/m

Patients underwent the following surgeries from October 2001 to September 2004 at hospitals participating in the Patient Safety in Surgery Study of the National Surgical Quality Improvement Program: 29 esophagectomy, 223 gastrectomy, 554 hepatectomy, 699 pancreatectomy, and 753 low anterior resection/proctectomy.

The risk of postoperative death was greatest at the extremes of BMI class, with a 30-day mortality rate of 9% among underweight patients, 2% among normal weight, 2.1% among overweight, 0.84% among obese I, 0.73% among obese II, and 3.8% among obese III, Dr. Mullen said.

In a multivariate analysis that examined 97 variables, the risk of postoperative death was significantly higher for underweight patients, with an odds ratio of 5.24, compared with patients stratified as normal (OR 1.00), overweight (OR 1.06), obese I (OR 0.61), obese II (OR 0.45), and obese III (OR 1.67).

The 30-day morbidity rate was 22% (OR 0.94), 23% (OR 1.00), 26% (OR 1.22), 29% (OR 1.42), 30% (OR 1.42), and 33% (OR 1.50), respectively, Dr. Mullen said.

There was a progressive and significant increase in minor complications, which included only urinary tract and wound infections, with increasing BMI class: underweight 9%, normal weight 13%, overweight 14%, obese I 17%, obese II 18%, and obese III 25%, said Dr. Mullen, who received no funding for the study and reported no conflicts of interest.

There were no differences among the BMI classes in total operative time or number of patients returning to the operating room.

Obesity has long been considered a potential risk factor for poor surgical outcomes, yet the published data are conflicting, Dr. Mullen said. The few studies that have examined outcomes after major cancer surgery have shown only an increased incidence of intra-abdominal and superficial wound infections. They were limited, however, by small numbers, retrospective design, limited patient follow-up, heterogeneity, and types of procedures studied, he said.

CHICAGO — Obesity was not a risk factor for postoperative mortality or major complications in patients undergoing major intra-abdominal cancer surgery, according to data from a prospective, multicenter risk-adjusted cohort study of 2,258 patients.

As observed in previous studies, however, obesity was a risk factor for minor complications, primarily wound infections.

Importantly, patients who were underweight had a fivefold increased risk of postoperative death, perhaps as a consequence of their underlying poor nutritional status or impaired immunity, investigator Dr. John T. Mullen reported at a symposium sponsored by the Society of Surgical Oncology.

“The prejudice that overweight and obese patients are at increased risk for serious adverse outcomes after major cancer surgery is not justified,” he said. “Paradoxically, overweight status and mild obesity status may be protective of low mortality.”

This possible “obesity paradox” warrants further study, whereas underweight patients might benefit from perioperative nutritional supplementation to mitigate their increased risk of mortality, Dr. Mullen, a surgical oncologist at Beth Israel Deaconess Medical Center, Boston, and his associates concluded.

Underweight patients in the study were significantly more likely to have lost more than 10% of their body weight in the 6 months prior to surgery and to have a lower mean preoperative serum albumin level than other patients have.

Using National Institutes of Health-defined body mass index (BMI) classes, 55 patients were stratified as underweight (BMI up to 18.5 kg/m

Patients underwent the following surgeries from October 2001 to September 2004 at hospitals participating in the Patient Safety in Surgery Study of the National Surgical Quality Improvement Program: 29 esophagectomy, 223 gastrectomy, 554 hepatectomy, 699 pancreatectomy, and 753 low anterior resection/proctectomy.

The risk of postoperative death was greatest at the extremes of BMI class, with a 30-day mortality rate of 9% among underweight patients, 2% among normal weight, 2.1% among overweight, 0.84% among obese I, 0.73% among obese II, and 3.8% among obese III, Dr. Mullen said.

In a multivariate analysis that examined 97 variables, the risk of postoperative death was significantly higher for underweight patients, with an odds ratio of 5.24, compared with patients stratified as normal (OR 1.00), overweight (OR 1.06), obese I (OR 0.61), obese II (OR 0.45), and obese III (OR 1.67).

The 30-day morbidity rate was 22% (OR 0.94), 23% (OR 1.00), 26% (OR 1.22), 29% (OR 1.42), 30% (OR 1.42), and 33% (OR 1.50), respectively, Dr. Mullen said.

There was a progressive and significant increase in minor complications, which included only urinary tract and wound infections, with increasing BMI class: underweight 9%, normal weight 13%, overweight 14%, obese I 17%, obese II 18%, and obese III 25%, said Dr. Mullen, who received no funding for the study and reported no conflicts of interest.

There were no differences among the BMI classes in total operative time or number of patients returning to the operating room.

Obesity has long been considered a potential risk factor for poor surgical outcomes, yet the published data are conflicting, Dr. Mullen said. The few studies that have examined outcomes after major cancer surgery have shown only an increased incidence of intra-abdominal and superficial wound infections. They were limited, however, by small numbers, retrospective design, limited patient follow-up, heterogeneity, and types of procedures studied, he said.

CHICAGO — Obesity was not a risk factor for postoperative mortality or major complications in patients undergoing major intra-abdominal cancer surgery, according to data from a prospective, multicenter risk-adjusted cohort study of 2,258 patients.

As observed in previous studies, however, obesity was a risk factor for minor complications, primarily wound infections.

Importantly, patients who were underweight had a fivefold increased risk of postoperative death, perhaps as a consequence of their underlying poor nutritional status or impaired immunity, investigator Dr. John T. Mullen reported at a symposium sponsored by the Society of Surgical Oncology.

“The prejudice that overweight and obese patients are at increased risk for serious adverse outcomes after major cancer surgery is not justified,” he said. “Paradoxically, overweight status and mild obesity status may be protective of low mortality.”

This possible “obesity paradox” warrants further study, whereas underweight patients might benefit from perioperative nutritional supplementation to mitigate their increased risk of mortality, Dr. Mullen, a surgical oncologist at Beth Israel Deaconess Medical Center, Boston, and his associates concluded.

Underweight patients in the study were significantly more likely to have lost more than 10% of their body weight in the 6 months prior to surgery and to have a lower mean preoperative serum albumin level than other patients have.

Using National Institutes of Health-defined body mass index (BMI) classes, 55 patients were stratified as underweight (BMI up to 18.5 kg/m

Patients underwent the following surgeries from October 2001 to September 2004 at hospitals participating in the Patient Safety in Surgery Study of the National Surgical Quality Improvement Program: 29 esophagectomy, 223 gastrectomy, 554 hepatectomy, 699 pancreatectomy, and 753 low anterior resection/proctectomy.

The risk of postoperative death was greatest at the extremes of BMI class, with a 30-day mortality rate of 9% among underweight patients, 2% among normal weight, 2.1% among overweight, 0.84% among obese I, 0.73% among obese II, and 3.8% among obese III, Dr. Mullen said.

In a multivariate analysis that examined 97 variables, the risk of postoperative death was significantly higher for underweight patients, with an odds ratio of 5.24, compared with patients stratified as normal (OR 1.00), overweight (OR 1.06), obese I (OR 0.61), obese II (OR 0.45), and obese III (OR 1.67).

The 30-day morbidity rate was 22% (OR 0.94), 23% (OR 1.00), 26% (OR 1.22), 29% (OR 1.42), 30% (OR 1.42), and 33% (OR 1.50), respectively, Dr. Mullen said.

There was a progressive and significant increase in minor complications, which included only urinary tract and wound infections, with increasing BMI class: underweight 9%, normal weight 13%, overweight 14%, obese I 17%, obese II 18%, and obese III 25%, said Dr. Mullen, who received no funding for the study and reported no conflicts of interest.

There were no differences among the BMI classes in total operative time or number of patients returning to the operating room.

Obesity has long been considered a potential risk factor for poor surgical outcomes, yet the published data are conflicting, Dr. Mullen said. The few studies that have examined outcomes after major cancer surgery have shown only an increased incidence of intra-abdominal and superficial wound infections. They were limited, however, by small numbers, retrospective design, limited patient follow-up, heterogeneity, and types of procedures studied, he said.