Sherry Boschert

SAN FRANCISCO — The first step in preparing a diabetic woman for pregnancy is noticing that she has diabetes before she conceives.

Women with type 2 diabetes often don't get diagnosed until pregnancy, by which time it's too late to reduce the risk of congenital anomalies through better glycemic control, Dr. Ingrid Block said at a meeting on diabetes and endocrinology sponsored by the University of California, San Francisco.

Congenital anomalies in infants of diabetic mothers occur as early as 5 weeks after the last menstrual period (for caudal regression) and as late as 8 weeks after the last period (for cardiac anomalies).

“If you don't sit down with that patient and ensure that she plans her pregnancy and that she has good glycemic control before conception, you run the risk that she'll find out she's 8 weeks pregnant and she has missed the opportunity” to avoid these congenital anomalies, said Dr. Block, of the university.

With any new female patients, pay attention to their obstetric histories, she urged. If a nondiabetic woman has delivered a large baby or had gestational diabetes, she's at increased risk for developing type 2 diabetes and should be screened for it periodically.

Congenital anomalies occur in 6%–10% of pregnancies among diabetic women with uncontrolled hyperglycemia, compared with an incidence of 2% in nondiabetic women. Emphasize effective contraception until diabetes patients achieve stable glycemia, Dr. Block said.

Preconception counseling and care should help women optimize glycemic control before pregnancy, which significantly reduces the risks of anomalies and fetal death, studies have shown. Women with type 2 diabetes should transition before conception from managing their diabetes using diet alone or oral therapies to using insulin, she added.

Identification and treatment of long-term complications of diabetes will give physicians an opportunity to warn some patients about difficult or nonviable pregnancies.

SAN FRANCISCO — The first step in preparing a diabetic woman for pregnancy is noticing that she has diabetes before she conceives.

Women with type 2 diabetes often don't get diagnosed until pregnancy, by which time it's too late to reduce the risk of congenital anomalies through better glycemic control, Dr. Ingrid Block said at a meeting on diabetes and endocrinology sponsored by the University of California, San Francisco.

Congenital anomalies in infants of diabetic mothers occur as early as 5 weeks after the last menstrual period (for caudal regression) and as late as 8 weeks after the last period (for cardiac anomalies).

“If you don't sit down with that patient and ensure that she plans her pregnancy and that she has good glycemic control before conception, you run the risk that she'll find out she's 8 weeks pregnant and she has missed the opportunity” to avoid these congenital anomalies, said Dr. Block, of the university.

With any new female patients, pay attention to their obstetric histories, she urged. If a nondiabetic woman has delivered a large baby or had gestational diabetes, she's at increased risk for developing type 2 diabetes and should be screened for it periodically.

Congenital anomalies occur in 6%–10% of pregnancies among diabetic women with uncontrolled hyperglycemia, compared with an incidence of 2% in nondiabetic women. Emphasize effective contraception until diabetes patients achieve stable glycemia, Dr. Block said.

Preconception counseling and care should help women optimize glycemic control before pregnancy, which significantly reduces the risks of anomalies and fetal death, studies have shown. Women with type 2 diabetes should transition before conception from managing their diabetes using diet alone or oral therapies to using insulin, she added.

Identification and treatment of long-term complications of diabetes will give physicians an opportunity to warn some patients about difficult or nonviable pregnancies.

SAN FRANCISCO — The first step in preparing a diabetic woman for pregnancy is noticing that she has diabetes before she conceives.

Women with type 2 diabetes often don't get diagnosed until pregnancy, by which time it's too late to reduce the risk of congenital anomalies through better glycemic control, Dr. Ingrid Block said at a meeting on diabetes and endocrinology sponsored by the University of California, San Francisco.

Congenital anomalies in infants of diabetic mothers occur as early as 5 weeks after the last menstrual period (for caudal regression) and as late as 8 weeks after the last period (for cardiac anomalies).

“If you don't sit down with that patient and ensure that she plans her pregnancy and that she has good glycemic control before conception, you run the risk that she'll find out she's 8 weeks pregnant and she has missed the opportunity” to avoid these congenital anomalies, said Dr. Block, of the university.

With any new female patients, pay attention to their obstetric histories, she urged. If a nondiabetic woman has delivered a large baby or had gestational diabetes, she's at increased risk for developing type 2 diabetes and should be screened for it periodically.

Congenital anomalies occur in 6%–10% of pregnancies among diabetic women with uncontrolled hyperglycemia, compared with an incidence of 2% in nondiabetic women. Emphasize effective contraception until diabetes patients achieve stable glycemia, Dr. Block said.

Preconception counseling and care should help women optimize glycemic control before pregnancy, which significantly reduces the risks of anomalies and fetal death, studies have shown. Women with type 2 diabetes should transition before conception from managing their diabetes using diet alone or oral therapies to using insulin, she added.

Identification and treatment of long-term complications of diabetes will give physicians an opportunity to warn some patients about difficult or nonviable pregnancies.

SAN FRANCISCO — Reducing dietary carbohydrates can improve atherogenic dyslipidemia, even in the absence of weight loss, Dr. Ronald M. Krauss said at a meeting on diabetes and endocrinology sponsored by the University of California, San Francisco.

Weight loss also improves dyslipidemia, mainly in people who have not already limited carbohydrates. For patients already restricting carbohydrates, weight loss offers little additional benefit to their lipid profiles, said Dr. Krauss, director of atherosclerosis research at Children's Hospital Oakland (Calif.) Research Institute.

To take the reduced-carbohydrate path, focus on avoiding high-glycemic starches and fructose, he advised.

He and his associates randomized 178 otherwise healthy overweight or obese men to one of four diets: one based on standard dietary recommendations (made up of 54% carbohydrates, 30% fat, and 16% protein), a diet with moderate carbohydrate reduction (to 39%), or one of two low (26%)-carbohydrate diets. To keep calorie levels the same initially, the researchers increased protein intake to 29% of the reduced-carbohydrate diet, and in the lowest carbohydrate diets increased either saturated or monounsaturated fat intake.

After 3 weeks (“which is enough to stabilize lipids with no weight change,” he said), they found a linear relationship between greater carbohydrate restriction and a change in the type of LDL cholesterol (Am. J. Clin. Nutr. 2006;83:1025–31). Carbohydrate restriction converted men from phenotype B individuals (who had dense, small-diameter LDL particles that confer higher atherogenic risk) to phenotype A (with medium-to-large-diameter LDL that's less risky).

Investigators then restricted calories, and patients in all groups lost similar amounts of weight. The type of diet “didn't make any difference as long as they learned to eat less,” he said. Lipid levels improved with weight loss, but less so in the low-carbohydrate groups that already had shown improvements.

The only significant reductions in LDL levels were seen with the low-carbohydrate, low-saturated-fat diet. “This is certainly the most effective diet in terms of LDL lowering that we've seen just by manipulating fat,” he added.

The reductions in small LDL particles from lowering carbohydrate intake were independent of saturated fat intake. Higher saturated fat intake did not attenuate the lipid benefits of lowering carbohydrates. Saturated fat intake “doesn't make it any worse. That's provocative, but that's what we found,” he said.

Previous studies of the lipid effects of low-carbohydrate diets didn't control for the effects of weight loss.

“You can get there either way. If you want to get the optimal result, you can either lose weight or you can drop carbohydrates. If you drop carbohydrates, it may not be as critical how much weight you lose,” Dr. Krauss said.

Other studies are attempting to replicate the findings. A recent study randomized dyslipidemic patients to one of four diets for 1 year: the severely low-carbohydrate Atkins diet, the more moderately carbohydrate-restricted Zone diet, a diet based on standard dietary recommendations, or the low-fat, high-carbohydrate Ornish diet.

The four groups had similar success in losing weight. Patients on the Atkins diet had somewhat better changes in body mass index, compared with the other groups, and profoundly better effects on lipid profiles, notably increases in HDL cholesterol and decreases in triglyceride levels (JAMA 2007;297:969–77).

Changes in the different types of LDL cholesterol were not measured in this cohort, but “there's just no doubt this would correspond to the same sort of changes we had seen,” Dr. Krauss said.

The most effective diet probably will be one that patients are able to maintain, he added. “In the end, that will be the biggest test of whether or not this works.”

To take the reduced-carbohydrate path, focus on avoiding high-glycemic starches and fructose. DR. KRAUSS

SAN FRANCISCO — Reducing dietary carbohydrates can improve atherogenic dyslipidemia, even in the absence of weight loss, Dr. Ronald M. Krauss said at a meeting on diabetes and endocrinology sponsored by the University of California, San Francisco.

Weight loss also improves dyslipidemia, mainly in people who have not already limited carbohydrates. For patients already restricting carbohydrates, weight loss offers little additional benefit to their lipid profiles, said Dr. Krauss, director of atherosclerosis research at Children's Hospital Oakland (Calif.) Research Institute.

To take the reduced-carbohydrate path, focus on avoiding high-glycemic starches and fructose, he advised.

He and his associates randomized 178 otherwise healthy overweight or obese men to one of four diets: one based on standard dietary recommendations (made up of 54% carbohydrates, 30% fat, and 16% protein), a diet with moderate carbohydrate reduction (to 39%), or one of two low (26%)-carbohydrate diets. To keep calorie levels the same initially, the researchers increased protein intake to 29% of the reduced-carbohydrate diet, and in the lowest carbohydrate diets increased either saturated or monounsaturated fat intake.

After 3 weeks (“which is enough to stabilize lipids with no weight change,” he said), they found a linear relationship between greater carbohydrate restriction and a change in the type of LDL cholesterol (Am. J. Clin. Nutr. 2006;83:1025–31). Carbohydrate restriction converted men from phenotype B individuals (who had dense, small-diameter LDL particles that confer higher atherogenic risk) to phenotype A (with medium-to-large-diameter LDL that's less risky).

Investigators then restricted calories, and patients in all groups lost similar amounts of weight. The type of diet “didn't make any difference as long as they learned to eat less,” he said. Lipid levels improved with weight loss, but less so in the low-carbohydrate groups that already had shown improvements.

The only significant reductions in LDL levels were seen with the low-carbohydrate, low-saturated-fat diet. “This is certainly the most effective diet in terms of LDL lowering that we've seen just by manipulating fat,” he added.

The reductions in small LDL particles from lowering carbohydrate intake were independent of saturated fat intake. Higher saturated fat intake did not attenuate the lipid benefits of lowering carbohydrates. Saturated fat intake “doesn't make it any worse. That's provocative, but that's what we found,” he said.

Previous studies of the lipid effects of low-carbohydrate diets didn't control for the effects of weight loss.

“You can get there either way. If you want to get the optimal result, you can either lose weight or you can drop carbohydrates. If you drop carbohydrates, it may not be as critical how much weight you lose,” Dr. Krauss said.

Other studies are attempting to replicate the findings. A recent study randomized dyslipidemic patients to one of four diets for 1 year: the severely low-carbohydrate Atkins diet, the more moderately carbohydrate-restricted Zone diet, a diet based on standard dietary recommendations, or the low-fat, high-carbohydrate Ornish diet.

The four groups had similar success in losing weight. Patients on the Atkins diet had somewhat better changes in body mass index, compared with the other groups, and profoundly better effects on lipid profiles, notably increases in HDL cholesterol and decreases in triglyceride levels (JAMA 2007;297:969–77).

Changes in the different types of LDL cholesterol were not measured in this cohort, but “there's just no doubt this would correspond to the same sort of changes we had seen,” Dr. Krauss said.

The most effective diet probably will be one that patients are able to maintain, he added. “In the end, that will be the biggest test of whether or not this works.”

To take the reduced-carbohydrate path, focus on avoiding high-glycemic starches and fructose. DR. KRAUSS

SAN FRANCISCO — Reducing dietary carbohydrates can improve atherogenic dyslipidemia, even in the absence of weight loss, Dr. Ronald M. Krauss said at a meeting on diabetes and endocrinology sponsored by the University of California, San Francisco.

Weight loss also improves dyslipidemia, mainly in people who have not already limited carbohydrates. For patients already restricting carbohydrates, weight loss offers little additional benefit to their lipid profiles, said Dr. Krauss, director of atherosclerosis research at Children's Hospital Oakland (Calif.) Research Institute.

To take the reduced-carbohydrate path, focus on avoiding high-glycemic starches and fructose, he advised.

He and his associates randomized 178 otherwise healthy overweight or obese men to one of four diets: one based on standard dietary recommendations (made up of 54% carbohydrates, 30% fat, and 16% protein), a diet with moderate carbohydrate reduction (to 39%), or one of two low (26%)-carbohydrate diets. To keep calorie levels the same initially, the researchers increased protein intake to 29% of the reduced-carbohydrate diet, and in the lowest carbohydrate diets increased either saturated or monounsaturated fat intake.

After 3 weeks (“which is enough to stabilize lipids with no weight change,” he said), they found a linear relationship between greater carbohydrate restriction and a change in the type of LDL cholesterol (Am. J. Clin. Nutr. 2006;83:1025–31). Carbohydrate restriction converted men from phenotype B individuals (who had dense, small-diameter LDL particles that confer higher atherogenic risk) to phenotype A (with medium-to-large-diameter LDL that's less risky).

Investigators then restricted calories, and patients in all groups lost similar amounts of weight. The type of diet “didn't make any difference as long as they learned to eat less,” he said. Lipid levels improved with weight loss, but less so in the low-carbohydrate groups that already had shown improvements.

The only significant reductions in LDL levels were seen with the low-carbohydrate, low-saturated-fat diet. “This is certainly the most effective diet in terms of LDL lowering that we've seen just by manipulating fat,” he added.

The reductions in small LDL particles from lowering carbohydrate intake were independent of saturated fat intake. Higher saturated fat intake did not attenuate the lipid benefits of lowering carbohydrates. Saturated fat intake “doesn't make it any worse. That's provocative, but that's what we found,” he said.

Previous studies of the lipid effects of low-carbohydrate diets didn't control for the effects of weight loss.

“You can get there either way. If you want to get the optimal result, you can either lose weight or you can drop carbohydrates. If you drop carbohydrates, it may not be as critical how much weight you lose,” Dr. Krauss said.

Other studies are attempting to replicate the findings. A recent study randomized dyslipidemic patients to one of four diets for 1 year: the severely low-carbohydrate Atkins diet, the more moderately carbohydrate-restricted Zone diet, a diet based on standard dietary recommendations, or the low-fat, high-carbohydrate Ornish diet.

The four groups had similar success in losing weight. Patients on the Atkins diet had somewhat better changes in body mass index, compared with the other groups, and profoundly better effects on lipid profiles, notably increases in HDL cholesterol and decreases in triglyceride levels (JAMA 2007;297:969–77).

Changes in the different types of LDL cholesterol were not measured in this cohort, but “there's just no doubt this would correspond to the same sort of changes we had seen,” Dr. Krauss said.

The most effective diet probably will be one that patients are able to maintain, he added. “In the end, that will be the biggest test of whether or not this works.”

To take the reduced-carbohydrate path, focus on avoiding high-glycemic starches and fructose. DR. KRAUSS

SAN FRANCISCO — Pregnant women with type 1 diabetes mellitus in a retrospective cohort study were more likely to improve glycemic control and less likely to deliver by cesarean section if they used insulin pumps rather than self-injections of insulin, according to Dr. Yvonne W. Cheng.

Among the 60 women in the pump group, 25% had hemoglobin A_1c (HbA_1c) values below 6%, compared with 13% of the 628 women in the injection group of a retrospective cohort study, she reported in a poster presentation at the annual meeting of the Society for Maternal-Fetal Medicine.

Half of women in the pump group delivered by C-section, compared with a 63% C-section rate in the injection group, said Dr. Cheng of the University of California, San Francisco, and her associates.

After controlling for the effects of maternal age, parity, ethnicity, body mass index, gestational weight gain, and gestational age at enrollment in the California Diabetes and Pregnancy Program, the study reported women in the pump group were three times as likely to have HbA_1c values below 6% and were half as likely to have a C-section, compared with the injection group.

The conclusions support results from one previous study in 2004 that found improved glycemic control with use of an insulin pump instead of injections by pregnant women with type 1 diabetes mellitus. Three other studies in 1988, 2000, and 2005 found no significant differences in results among groups, she noted. All the previous studies were smaller than the present study, with only 11–36 patients in the pump groups.

The current study also found that women in the pump group were more likely to be white, to speak English as their primary language, and to have a higher education level than did women in the injection group.

“We need to address the disparity of insulin pump use in type 1 diabetes mellitus patients of different socioeconomic and racial/ethnic groups,” the authors commented.

“In nonpregnant diabetics, most people are switching over to pumps” because studies have shown better glycemic control, Dr. Cheng said in an interview. But to be candidates for insulin pumps, women must be able to count carbohydrates and operate the machine, as well as program it.

“It's a very select group of women,” she said.

Dr. Cheng has no association with companies that make insulin pumps or injection products.

SAN FRANCISCO — Pregnant women with type 1 diabetes mellitus in a retrospective cohort study were more likely to improve glycemic control and less likely to deliver by cesarean section if they used insulin pumps rather than self-injections of insulin, according to Dr. Yvonne W. Cheng.

Among the 60 women in the pump group, 25% had hemoglobin A_1c (HbA_1c) values below 6%, compared with 13% of the 628 women in the injection group of a retrospective cohort study, she reported in a poster presentation at the annual meeting of the Society for Maternal-Fetal Medicine.

Half of women in the pump group delivered by C-section, compared with a 63% C-section rate in the injection group, said Dr. Cheng of the University of California, San Francisco, and her associates.

After controlling for the effects of maternal age, parity, ethnicity, body mass index, gestational weight gain, and gestational age at enrollment in the California Diabetes and Pregnancy Program, the study reported women in the pump group were three times as likely to have HbA_1c values below 6% and were half as likely to have a C-section, compared with the injection group.

The conclusions support results from one previous study in 2004 that found improved glycemic control with use of an insulin pump instead of injections by pregnant women with type 1 diabetes mellitus. Three other studies in 1988, 2000, and 2005 found no significant differences in results among groups, she noted. All the previous studies were smaller than the present study, with only 11–36 patients in the pump groups.

The current study also found that women in the pump group were more likely to be white, to speak English as their primary language, and to have a higher education level than did women in the injection group.

“We need to address the disparity of insulin pump use in type 1 diabetes mellitus patients of different socioeconomic and racial/ethnic groups,” the authors commented.

“In nonpregnant diabetics, most people are switching over to pumps” because studies have shown better glycemic control, Dr. Cheng said in an interview. But to be candidates for insulin pumps, women must be able to count carbohydrates and operate the machine, as well as program it.

“It's a very select group of women,” she said.

Dr. Cheng has no association with companies that make insulin pumps or injection products.

SAN FRANCISCO — Pregnant women with type 1 diabetes mellitus in a retrospective cohort study were more likely to improve glycemic control and less likely to deliver by cesarean section if they used insulin pumps rather than self-injections of insulin, according to Dr. Yvonne W. Cheng.

Among the 60 women in the pump group, 25% had hemoglobin A_1c (HbA_1c) values below 6%, compared with 13% of the 628 women in the injection group of a retrospective cohort study, she reported in a poster presentation at the annual meeting of the Society for Maternal-Fetal Medicine.

Half of women in the pump group delivered by C-section, compared with a 63% C-section rate in the injection group, said Dr. Cheng of the University of California, San Francisco, and her associates.

After controlling for the effects of maternal age, parity, ethnicity, body mass index, gestational weight gain, and gestational age at enrollment in the California Diabetes and Pregnancy Program, the study reported women in the pump group were three times as likely to have HbA_1c values below 6% and were half as likely to have a C-section, compared with the injection group.

The conclusions support results from one previous study in 2004 that found improved glycemic control with use of an insulin pump instead of injections by pregnant women with type 1 diabetes mellitus. Three other studies in 1988, 2000, and 2005 found no significant differences in results among groups, she noted. All the previous studies were smaller than the present study, with only 11–36 patients in the pump groups.

The current study also found that women in the pump group were more likely to be white, to speak English as their primary language, and to have a higher education level than did women in the injection group.

“We need to address the disparity of insulin pump use in type 1 diabetes mellitus patients of different socioeconomic and racial/ethnic groups,” the authors commented.

“In nonpregnant diabetics, most people are switching over to pumps” because studies have shown better glycemic control, Dr. Cheng said in an interview. But to be candidates for insulin pumps, women must be able to count carbohydrates and operate the machine, as well as program it.

“It's a very select group of women,” she said.

Dr. Cheng has no association with companies that make insulin pumps or injection products.

SAN FRANCISCO — Pregnant women with type 1 diabetes mellitus were more likely to improve glycemic control and less likely to deliver by cesarean section if they used insulin pumps rather than self-injections of insulin, Dr. Yvonne W. Cheng said.

Among 60 women in the pump group, 25% had hemoglobin A1c (HbA1c) values below 6%, compared with 13% of 628 women in the injection group of a retrospective cohort study, she reported in a poster presentation at the annual meeting of the Society for Maternal-Fetal Medicine.

Half of women in the pump group delivered by C-section, compared with a 63% C-section rate in the injection group, said Dr. Cheng of the University of California, San Francisco, and her associates.

After controlling for the effects of maternal age, parity, ethnicity, body mass index, gestational weight gain, and gestational age at enrollment in the California Diabetes and Pregnancy Program, women in the pump group were three times as likely to have HbA1c values below 6% and were half as likely to have a C-section, compared with the injection group.

The conclusions support results from one previous study in 2004 that found improved glycemic control with use of an insulin pump instead of injections by pregnant women with type 1 diabetes mellitus.

Three other studies in 1988, 2000, and 2005 found no significant differences in results among groups, she noted.

All the previous studies were smaller than the present study, with only 11-36 patients in the pump groups.

The current study also found that women in the pump group were more likely to be white, to speak English as their primary language, and to have a higher education level than did women in the injection group.

“We need to address the disparity of insulin pump use in type 1 diabetes mellitus patients of different socioeconomic and racial/ethnic groups,” the authors commented.

In terms of rate of preterm delivery, rate of large-for-gestational-age babies, or rate of admissions to intensive care nurseries, the researchers found that there were no differences between the pump and injections groups.

“In nonpregnant diabetics, most people are switching over to pumps” because studies have shown better glycemic control, Dr. Cheng said in an interview at her poster.

The pump provides continuous release of insulin, functioning more like the pancreas than do timed injections of insulin.

In order to be candidates for insulin pumps, women must be able to count carbohydrates, operate the machine, and program it.

“It's a very select group of women,” she said.

Dr. Cheng disclosed that she has no association with companies that make either insulin pumps or injection products.

ELSEVIER GLOBAL MEDICAL NEWS

SAN FRANCISCO — Pregnant women with type 1 diabetes mellitus were more likely to improve glycemic control and less likely to deliver by cesarean section if they used insulin pumps rather than self-injections of insulin, Dr. Yvonne W. Cheng said.

Among 60 women in the pump group, 25% had hemoglobin A1c (HbA1c) values below 6%, compared with 13% of 628 women in the injection group of a retrospective cohort study, she reported in a poster presentation at the annual meeting of the Society for Maternal-Fetal Medicine.

Half of women in the pump group delivered by C-section, compared with a 63% C-section rate in the injection group, said Dr. Cheng of the University of California, San Francisco, and her associates.

After controlling for the effects of maternal age, parity, ethnicity, body mass index, gestational weight gain, and gestational age at enrollment in the California Diabetes and Pregnancy Program, women in the pump group were three times as likely to have HbA1c values below 6% and were half as likely to have a C-section, compared with the injection group.

The conclusions support results from one previous study in 2004 that found improved glycemic control with use of an insulin pump instead of injections by pregnant women with type 1 diabetes mellitus.

Three other studies in 1988, 2000, and 2005 found no significant differences in results among groups, she noted.

All the previous studies were smaller than the present study, with only 11-36 patients in the pump groups.

The current study also found that women in the pump group were more likely to be white, to speak English as their primary language, and to have a higher education level than did women in the injection group.

“We need to address the disparity of insulin pump use in type 1 diabetes mellitus patients of different socioeconomic and racial/ethnic groups,” the authors commented.

In terms of rate of preterm delivery, rate of large-for-gestational-age babies, or rate of admissions to intensive care nurseries, the researchers found that there were no differences between the pump and injections groups.

“In nonpregnant diabetics, most people are switching over to pumps” because studies have shown better glycemic control, Dr. Cheng said in an interview at her poster.

The pump provides continuous release of insulin, functioning more like the pancreas than do timed injections of insulin.

In order to be candidates for insulin pumps, women must be able to count carbohydrates, operate the machine, and program it.

“It's a very select group of women,” she said.

Dr. Cheng disclosed that she has no association with companies that make either insulin pumps or injection products.

ELSEVIER GLOBAL MEDICAL NEWS

SAN FRANCISCO — Pregnant women with type 1 diabetes mellitus were more likely to improve glycemic control and less likely to deliver by cesarean section if they used insulin pumps rather than self-injections of insulin, Dr. Yvonne W. Cheng said.

Among 60 women in the pump group, 25% had hemoglobin A1c (HbA1c) values below 6%, compared with 13% of 628 women in the injection group of a retrospective cohort study, she reported in a poster presentation at the annual meeting of the Society for Maternal-Fetal Medicine.

Half of women in the pump group delivered by C-section, compared with a 63% C-section rate in the injection group, said Dr. Cheng of the University of California, San Francisco, and her associates.

After controlling for the effects of maternal age, parity, ethnicity, body mass index, gestational weight gain, and gestational age at enrollment in the California Diabetes and Pregnancy Program, women in the pump group were three times as likely to have HbA1c values below 6% and were half as likely to have a C-section, compared with the injection group.

The conclusions support results from one previous study in 2004 that found improved glycemic control with use of an insulin pump instead of injections by pregnant women with type 1 diabetes mellitus.

Three other studies in 1988, 2000, and 2005 found no significant differences in results among groups, she noted.

All the previous studies were smaller than the present study, with only 11-36 patients in the pump groups.

The current study also found that women in the pump group were more likely to be white, to speak English as their primary language, and to have a higher education level than did women in the injection group.

“We need to address the disparity of insulin pump use in type 1 diabetes mellitus patients of different socioeconomic and racial/ethnic groups,” the authors commented.

In terms of rate of preterm delivery, rate of large-for-gestational-age babies, or rate of admissions to intensive care nurseries, the researchers found that there were no differences between the pump and injections groups.

“In nonpregnant diabetics, most people are switching over to pumps” because studies have shown better glycemic control, Dr. Cheng said in an interview at her poster.

The pump provides continuous release of insulin, functioning more like the pancreas than do timed injections of insulin.

In order to be candidates for insulin pumps, women must be able to count carbohydrates, operate the machine, and program it.

“It's a very select group of women,” she said.

Dr. Cheng disclosed that she has no association with companies that make either insulin pumps or injection products.

ELSEVIER GLOBAL MEDICAL NEWS

SAN FRANCISCO — Delivery-related perinatal death rates aren't higher for second twins overall, but they more than double in twins delivered at term, British study results suggest.

Gestational age and birth order of twins made a statistically significant difference in an analysis of data from national registries in England, Northern Ireland, and Wales; the data were collected between 1994 and 2003.

The study included 1,501 cases of intrapartum stillbirth or neonatal death of the second twin but not the first, Dr. Gordon C. Smith reported at the annual meeting of the Society for Maternal-Fetal Medicine.

The risk of death due to intrapartum anoxia or trauma in second twins was three- to fourfold higher at term (at least 36 weeks' gestation), compared with earlier deliveries.

Similarly, risk was found to be fivefold higher in second twins delivered vaginally at term, compared with babies delivered earlier.

No statistically significant increase in risk was seen with cesarean deliveries at term, said Dr. Smith of the University of Cambridge, England.

“Occasionally, the risk of death may be reduced by planned cesarean section” of twins, he said.

One physician in the audience bemoaned the widespread use of cesarean section for delivery of fetuses in breech position and for many other indications.

“Now every second twin at term? Where are we going to stop?” he asked.

Dr. Smith suggested that physicians should try to balance the risks of cesarean section with the potential benefits for each patient in their counseling and management of pregnant women.

It has been well known that vaginal delivery of a second twin carries increased risks because of a number of specific complications like cord prolapse or placental abruption, but data have been mixed on whether the risk of neonatal death increases.

British studies in 2002 and 2005 found an increased risk of perinatal death for the second twin, but a U.S. analysis of 300,000 twin births found no association between birth order and risk of neonatal death.

The previous studies all had fundamental flaws in their analytical approaches that undermined their conclusions, Dr. Smith said.

The present study excluded perinatal deaths due to congenital anomalies.

SAN FRANCISCO — Delivery-related perinatal death rates aren't higher for second twins overall, but they more than double in twins delivered at term, British study results suggest.

Gestational age and birth order of twins made a statistically significant difference in an analysis of data from national registries in England, Northern Ireland, and Wales; the data were collected between 1994 and 2003.

The study included 1,501 cases of intrapartum stillbirth or neonatal death of the second twin but not the first, Dr. Gordon C. Smith reported at the annual meeting of the Society for Maternal-Fetal Medicine.

The risk of death due to intrapartum anoxia or trauma in second twins was three- to fourfold higher at term (at least 36 weeks' gestation), compared with earlier deliveries.

Similarly, risk was found to be fivefold higher in second twins delivered vaginally at term, compared with babies delivered earlier.

No statistically significant increase in risk was seen with cesarean deliveries at term, said Dr. Smith of the University of Cambridge, England.

“Occasionally, the risk of death may be reduced by planned cesarean section” of twins, he said.

One physician in the audience bemoaned the widespread use of cesarean section for delivery of fetuses in breech position and for many other indications.

“Now every second twin at term? Where are we going to stop?” he asked.

Dr. Smith suggested that physicians should try to balance the risks of cesarean section with the potential benefits for each patient in their counseling and management of pregnant women.

It has been well known that vaginal delivery of a second twin carries increased risks because of a number of specific complications like cord prolapse or placental abruption, but data have been mixed on whether the risk of neonatal death increases.

British studies in 2002 and 2005 found an increased risk of perinatal death for the second twin, but a U.S. analysis of 300,000 twin births found no association between birth order and risk of neonatal death.

The previous studies all had fundamental flaws in their analytical approaches that undermined their conclusions, Dr. Smith said.

The present study excluded perinatal deaths due to congenital anomalies.

SAN FRANCISCO — Delivery-related perinatal death rates aren't higher for second twins overall, but they more than double in twins delivered at term, British study results suggest.

Gestational age and birth order of twins made a statistically significant difference in an analysis of data from national registries in England, Northern Ireland, and Wales; the data were collected between 1994 and 2003.

The study included 1,501 cases of intrapartum stillbirth or neonatal death of the second twin but not the first, Dr. Gordon C. Smith reported at the annual meeting of the Society for Maternal-Fetal Medicine.

The risk of death due to intrapartum anoxia or trauma in second twins was three- to fourfold higher at term (at least 36 weeks' gestation), compared with earlier deliveries.

Similarly, risk was found to be fivefold higher in second twins delivered vaginally at term, compared with babies delivered earlier.

No statistically significant increase in risk was seen with cesarean deliveries at term, said Dr. Smith of the University of Cambridge, England.

“Occasionally, the risk of death may be reduced by planned cesarean section” of twins, he said.

One physician in the audience bemoaned the widespread use of cesarean section for delivery of fetuses in breech position and for many other indications.

“Now every second twin at term? Where are we going to stop?” he asked.

Dr. Smith suggested that physicians should try to balance the risks of cesarean section with the potential benefits for each patient in their counseling and management of pregnant women.

It has been well known that vaginal delivery of a second twin carries increased risks because of a number of specific complications like cord prolapse or placental abruption, but data have been mixed on whether the risk of neonatal death increases.

British studies in 2002 and 2005 found an increased risk of perinatal death for the second twin, but a U.S. analysis of 300,000 twin births found no association between birth order and risk of neonatal death.

The previous studies all had fundamental flaws in their analytical approaches that undermined their conclusions, Dr. Smith said.

The present study excluded perinatal deaths due to congenital anomalies.

SAN FRANCISCO — Evaluating persistent headache that presents more than 24 hours after delivery requires a stepwise, multidisciplinary approach, Dr. Caroline Stella said at the annual meeting of the Society for Maternal-Fetal Medicine.

Postpartum headache affects 11%–80% of women after delivery, and usually is benign. Little has been known about headaches that start more than a day after delivery. Dr. Stella of the University of Cincinnati and her associates retrospectively studied records for 95 women with severe headaches that started 25 hours to 32 days post partum and were unresponsive to usual doses of analgesics.

Approximately half of the women (47 patients) ultimately were diagnosed with tension-type or migraine headaches, and all responded to higher doses of analgesics or narcotics. Twenty-three women (24%) with headaches caused by preeclampsia or eclampsia were treated with magnesium sulfate or antihypertensives; one had a cerebral venous thrombosis, and one had a subarachnoid hemorrhage.

Fifteen women (16%) had a final diagnosis of spinal headache. All of them received an anesthesiology consultation and initially were treated with analgesics. Twelve patients eventually required a blood patch. Headaches caused by other abnormalities in 10 women (11%) were treated with anticonvulsants, anticoagulants, or a dopamine antagonist.

Of the 95 women, 22 (23%) underwent cerebral imaging studies. Indications for imaging included persistent visual changes in 14 patients, focal neurologic deficits in 8, refractory headache in 8, and other reasons.

Neurologic symptoms other than headache resolved in 20 of the 22 women. Two patients were left with residual facial droop. No patients died.

On the basis of the study's findings, Dr. Stella and her associates recommended the following algorithm for work-up of postpartum headache presenting after 24 hours.

If neurologic deficits are present, perform cerebral imaging right away. If no neurologic deficits are present in a normotensive patient without proteinuria, treat for presumed migraine or tension-type headache.

If that therapy doesn't work and the patient has a history of regional anesthesia placement, consider the diagnosis of postdural puncture headache. Give intravenous fluids and analgesia, and consider administering a blood patch. If the headache still doesn't respond to therapy, perform cerebral imaging.

Presume a diagnosis of preeclampsia or eclampsia in women with hypertension or proteinuria and administer magnesium sulfate and antihypertensive agents. If the headache is unresponsive, perform cerebral imaging.

Eclampsia increased the risk for cognitive dysfunction years later in a separate case-control study of 87 women presented in a poster session by Annet M. Aukes, a medical student at the University of Groningen (the Netherlands).

Six to 8 years after their eclampsia episodes, 29 women were matched by age and year of pregnancy with formerly preeclamptic women and normotensive controls. All completed a validated Cognitive Failures Questionnaire designed to assess the likelihood of committing a cognitive error in an everyday task.

The women with a history of eclampsia scored significantly worse, compared with preeclamptic women or controls. “It's very important that these women are treated adequately right away when they come in with headaches so we can prevent more eclamptic seizures,” Ms. Aukes said.

A hyperintense cortical and subcortical signal in the occipital lobes is consistentwith a common eclampsia lesion. Courtesy Dr. Caroline Stella

SAN FRANCISCO — Evaluating persistent headache that presents more than 24 hours after delivery requires a stepwise, multidisciplinary approach, Dr. Caroline Stella said at the annual meeting of the Society for Maternal-Fetal Medicine.

Postpartum headache affects 11%–80% of women after delivery, and usually is benign. Little has been known about headaches that start more than a day after delivery. Dr. Stella of the University of Cincinnati and her associates retrospectively studied records for 95 women with severe headaches that started 25 hours to 32 days post partum and were unresponsive to usual doses of analgesics.

Approximately half of the women (47 patients) ultimately were diagnosed with tension-type or migraine headaches, and all responded to higher doses of analgesics or narcotics. Twenty-three women (24%) with headaches caused by preeclampsia or eclampsia were treated with magnesium sulfate or antihypertensives; one had a cerebral venous thrombosis, and one had a subarachnoid hemorrhage.

Fifteen women (16%) had a final diagnosis of spinal headache. All of them received an anesthesiology consultation and initially were treated with analgesics. Twelve patients eventually required a blood patch. Headaches caused by other abnormalities in 10 women (11%) were treated with anticonvulsants, anticoagulants, or a dopamine antagonist.

Of the 95 women, 22 (23%) underwent cerebral imaging studies. Indications for imaging included persistent visual changes in 14 patients, focal neurologic deficits in 8, refractory headache in 8, and other reasons.

Neurologic symptoms other than headache resolved in 20 of the 22 women. Two patients were left with residual facial droop. No patients died.

On the basis of the study's findings, Dr. Stella and her associates recommended the following algorithm for work-up of postpartum headache presenting after 24 hours.

If neurologic deficits are present, perform cerebral imaging right away. If no neurologic deficits are present in a normotensive patient without proteinuria, treat for presumed migraine or tension-type headache.

If that therapy doesn't work and the patient has a history of regional anesthesia placement, consider the diagnosis of postdural puncture headache. Give intravenous fluids and analgesia, and consider administering a blood patch. If the headache still doesn't respond to therapy, perform cerebral imaging.

Presume a diagnosis of preeclampsia or eclampsia in women with hypertension or proteinuria and administer magnesium sulfate and antihypertensive agents. If the headache is unresponsive, perform cerebral imaging.

Eclampsia increased the risk for cognitive dysfunction years later in a separate case-control study of 87 women presented in a poster session by Annet M. Aukes, a medical student at the University of Groningen (the Netherlands).

Six to 8 years after their eclampsia episodes, 29 women were matched by age and year of pregnancy with formerly preeclamptic women and normotensive controls. All completed a validated Cognitive Failures Questionnaire designed to assess the likelihood of committing a cognitive error in an everyday task.

The women with a history of eclampsia scored significantly worse, compared with preeclamptic women or controls. “It's very important that these women are treated adequately right away when they come in with headaches so we can prevent more eclamptic seizures,” Ms. Aukes said.

A hyperintense cortical and subcortical signal in the occipital lobes is consistentwith a common eclampsia lesion. Courtesy Dr. Caroline Stella

SAN FRANCISCO — Evaluating persistent headache that presents more than 24 hours after delivery requires a stepwise, multidisciplinary approach, Dr. Caroline Stella said at the annual meeting of the Society for Maternal-Fetal Medicine.

Postpartum headache affects 11%–80% of women after delivery, and usually is benign. Little has been known about headaches that start more than a day after delivery. Dr. Stella of the University of Cincinnati and her associates retrospectively studied records for 95 women with severe headaches that started 25 hours to 32 days post partum and were unresponsive to usual doses of analgesics.

Approximately half of the women (47 patients) ultimately were diagnosed with tension-type or migraine headaches, and all responded to higher doses of analgesics or narcotics. Twenty-three women (24%) with headaches caused by preeclampsia or eclampsia were treated with magnesium sulfate or antihypertensives; one had a cerebral venous thrombosis, and one had a subarachnoid hemorrhage.

Fifteen women (16%) had a final diagnosis of spinal headache. All of them received an anesthesiology consultation and initially were treated with analgesics. Twelve patients eventually required a blood patch. Headaches caused by other abnormalities in 10 women (11%) were treated with anticonvulsants, anticoagulants, or a dopamine antagonist.

Of the 95 women, 22 (23%) underwent cerebral imaging studies. Indications for imaging included persistent visual changes in 14 patients, focal neurologic deficits in 8, refractory headache in 8, and other reasons.

Neurologic symptoms other than headache resolved in 20 of the 22 women. Two patients were left with residual facial droop. No patients died.

On the basis of the study's findings, Dr. Stella and her associates recommended the following algorithm for work-up of postpartum headache presenting after 24 hours.

If neurologic deficits are present, perform cerebral imaging right away. If no neurologic deficits are present in a normotensive patient without proteinuria, treat for presumed migraine or tension-type headache.

If that therapy doesn't work and the patient has a history of regional anesthesia placement, consider the diagnosis of postdural puncture headache. Give intravenous fluids and analgesia, and consider administering a blood patch. If the headache still doesn't respond to therapy, perform cerebral imaging.

Presume a diagnosis of preeclampsia or eclampsia in women with hypertension or proteinuria and administer magnesium sulfate and antihypertensive agents. If the headache is unresponsive, perform cerebral imaging.

Eclampsia increased the risk for cognitive dysfunction years later in a separate case-control study of 87 women presented in a poster session by Annet M. Aukes, a medical student at the University of Groningen (the Netherlands).

Six to 8 years after their eclampsia episodes, 29 women were matched by age and year of pregnancy with formerly preeclamptic women and normotensive controls. All completed a validated Cognitive Failures Questionnaire designed to assess the likelihood of committing a cognitive error in an everyday task.

The women with a history of eclampsia scored significantly worse, compared with preeclamptic women or controls. “It's very important that these women are treated adequately right away when they come in with headaches so we can prevent more eclamptic seizures,” Ms. Aukes said.

A hyperintense cortical and subcortical signal in the occipital lobes is consistentwith a common eclampsia lesion. Courtesy Dr. Caroline Stella

SAN FRANCISCO — Getting a protein-to-creatinine ratio was more helpful than using a urine dipstick to measure proteinuria in patients with suspected preeclampsia, according to a retrospective cohort study.

The protein to creatinine (P/C) ratio correlated strongly with a 24-hour urine protein measurement, which is the standard for quantifying protein. The P/C ratio had a 90% correlation with 24-hour urine protein measurements, compared with only a 58% correlation between the urine dipstick and 24-hour urine protein measurements, Jasmine Lai and associates reported in a poster presentation at the annual meeting of the Society for Maternal-Fetal Medicine.

Investigators analyzed data on 140 women with suspected preeclampsia who had both a dipstick and 24-hour protein measurement, 177 who had both a P/C ratio and 24-hour protein measurement, and 244 who had both a dipstick and P/C ratio. The different assays were performed within 48 hours of each other for each patient.

The P/C ratio was a more sensitive marker for proteinuria, with a sensitivity of 75%, compared with dipstick's sensitivity of 44%, reported Ms. Lai, a student at the University of California, San Francisco, who conducted the study while a summer fellow at the University of California, San Diego.

Now that getting a dipstick measurement has been encumbered by the Clinical Laboratory Improvement Amendments law, it's just as fast and efficient to get a P/C ratio, Dr. Douglas Woelkers, the primary investigator in the study, said in an interview.

“Nurses now can't do dipsticks in the [labor and delivery] setting. Our hospital requires that all dipsticks go down to the laboratory to be read by machine. Why not get the more accurate P/C ratio, because it takes the same time to get a result back, and it's the same expense compared with the dipstick?” said Dr. Woelkers of the University of California, San Diego.

The dipstick underestimated proteinuria 44%–48% of the time, he added. Patients with a false-negative dipstick and mild hypertension would be sent home “only to find out later on that they truly had the disease and we weren't intervening soon enough,” he said.

Dipstick measurements were significantly confounded by the method of collection and the presence of blood, squamous cells, white blood cells, or leukocyte esterase, the investigators found.

To confirm the superiority of the P/C ratio, the investigators analyzed adverse maternal or fetal outcomes from delivery records of 209 patients who had both a urine dipstick and a P/C ratio. They used a composite of three or more markers for severe disease, including thrombocytopenia, elevated liver function tests, high creatinine level, low Apgar score, low birth weight, and maternal hospitalization longer than 3 days.

The P/C ratio was more accurate than the dipstick in predicting adverse outcomes because it more accurately measured proteinuria, Dr. Woelkers said. The P/C ratio had a sensitivity of 50% for composite adverse outcomes and a specificity of 72%. Taking one or more dipstick measurements was 35% sensitive and 81% specific for predicting adverse outcomes. Taking two or more dipstick measurements was 24% sensitive and 83% specific for adverse outcomes. Dr. Woelkers' hospital has converted entirely to doing P/C ratios instead of dipsticks for patients with suspected preeclampsia.

Further research will be needed to see if it makes sense to switch from dipsticks to P/C ratios for patients with suspected preeclampsia in office settings, not just in hospitals, he added.

SAN FRANCISCO — Getting a protein-to-creatinine ratio was more helpful than using a urine dipstick to measure proteinuria in patients with suspected preeclampsia, according to a retrospective cohort study.

The protein to creatinine (P/C) ratio correlated strongly with a 24-hour urine protein measurement, which is the standard for quantifying protein. The P/C ratio had a 90% correlation with 24-hour urine protein measurements, compared with only a 58% correlation between the urine dipstick and 24-hour urine protein measurements, Jasmine Lai and associates reported in a poster presentation at the annual meeting of the Society for Maternal-Fetal Medicine.

Investigators analyzed data on 140 women with suspected preeclampsia who had both a dipstick and 24-hour protein measurement, 177 who had both a P/C ratio and 24-hour protein measurement, and 244 who had both a dipstick and P/C ratio. The different assays were performed within 48 hours of each other for each patient.

The P/C ratio was a more sensitive marker for proteinuria, with a sensitivity of 75%, compared with dipstick's sensitivity of 44%, reported Ms. Lai, a student at the University of California, San Francisco, who conducted the study while a summer fellow at the University of California, San Diego.

Now that getting a dipstick measurement has been encumbered by the Clinical Laboratory Improvement Amendments law, it's just as fast and efficient to get a P/C ratio, Dr. Douglas Woelkers, the primary investigator in the study, said in an interview.

“Nurses now can't do dipsticks in the [labor and delivery] setting. Our hospital requires that all dipsticks go down to the laboratory to be read by machine. Why not get the more accurate P/C ratio, because it takes the same time to get a result back, and it's the same expense compared with the dipstick?” said Dr. Woelkers of the University of California, San Diego.

The dipstick underestimated proteinuria 44%–48% of the time, he added. Patients with a false-negative dipstick and mild hypertension would be sent home “only to find out later on that they truly had the disease and we weren't intervening soon enough,” he said.

Dipstick measurements were significantly confounded by the method of collection and the presence of blood, squamous cells, white blood cells, or leukocyte esterase, the investigators found.

To confirm the superiority of the P/C ratio, the investigators analyzed adverse maternal or fetal outcomes from delivery records of 209 patients who had both a urine dipstick and a P/C ratio. They used a composite of three or more markers for severe disease, including thrombocytopenia, elevated liver function tests, high creatinine level, low Apgar score, low birth weight, and maternal hospitalization longer than 3 days.

The P/C ratio was more accurate than the dipstick in predicting adverse outcomes because it more accurately measured proteinuria, Dr. Woelkers said. The P/C ratio had a sensitivity of 50% for composite adverse outcomes and a specificity of 72%. Taking one or more dipstick measurements was 35% sensitive and 81% specific for predicting adverse outcomes. Taking two or more dipstick measurements was 24% sensitive and 83% specific for adverse outcomes. Dr. Woelkers' hospital has converted entirely to doing P/C ratios instead of dipsticks for patients with suspected preeclampsia.

Further research will be needed to see if it makes sense to switch from dipsticks to P/C ratios for patients with suspected preeclampsia in office settings, not just in hospitals, he added.

SAN FRANCISCO — Getting a protein-to-creatinine ratio was more helpful than using a urine dipstick to measure proteinuria in patients with suspected preeclampsia, according to a retrospective cohort study.

The protein to creatinine (P/C) ratio correlated strongly with a 24-hour urine protein measurement, which is the standard for quantifying protein. The P/C ratio had a 90% correlation with 24-hour urine protein measurements, compared with only a 58% correlation between the urine dipstick and 24-hour urine protein measurements, Jasmine Lai and associates reported in a poster presentation at the annual meeting of the Society for Maternal-Fetal Medicine.

Investigators analyzed data on 140 women with suspected preeclampsia who had both a dipstick and 24-hour protein measurement, 177 who had both a P/C ratio and 24-hour protein measurement, and 244 who had both a dipstick and P/C ratio. The different assays were performed within 48 hours of each other for each patient.

The P/C ratio was a more sensitive marker for proteinuria, with a sensitivity of 75%, compared with dipstick's sensitivity of 44%, reported Ms. Lai, a student at the University of California, San Francisco, who conducted the study while a summer fellow at the University of California, San Diego.

Now that getting a dipstick measurement has been encumbered by the Clinical Laboratory Improvement Amendments law, it's just as fast and efficient to get a P/C ratio, Dr. Douglas Woelkers, the primary investigator in the study, said in an interview.

“Nurses now can't do dipsticks in the [labor and delivery] setting. Our hospital requires that all dipsticks go down to the laboratory to be read by machine. Why not get the more accurate P/C ratio, because it takes the same time to get a result back, and it's the same expense compared with the dipstick?” said Dr. Woelkers of the University of California, San Diego.

The dipstick underestimated proteinuria 44%–48% of the time, he added. Patients with a false-negative dipstick and mild hypertension would be sent home “only to find out later on that they truly had the disease and we weren't intervening soon enough,” he said.

Dipstick measurements were significantly confounded by the method of collection and the presence of blood, squamous cells, white blood cells, or leukocyte esterase, the investigators found.

To confirm the superiority of the P/C ratio, the investigators analyzed adverse maternal or fetal outcomes from delivery records of 209 patients who had both a urine dipstick and a P/C ratio. They used a composite of three or more markers for severe disease, including thrombocytopenia, elevated liver function tests, high creatinine level, low Apgar score, low birth weight, and maternal hospitalization longer than 3 days.

The P/C ratio was more accurate than the dipstick in predicting adverse outcomes because it more accurately measured proteinuria, Dr. Woelkers said. The P/C ratio had a sensitivity of 50% for composite adverse outcomes and a specificity of 72%. Taking one or more dipstick measurements was 35% sensitive and 81% specific for predicting adverse outcomes. Taking two or more dipstick measurements was 24% sensitive and 83% specific for adverse outcomes. Dr. Woelkers' hospital has converted entirely to doing P/C ratios instead of dipsticks for patients with suspected preeclampsia.

Further research will be needed to see if it makes sense to switch from dipsticks to P/C ratios for patients with suspected preeclampsia in office settings, not just in hospitals, he added.

SAN FRANCISCO — Evaluating persistent headache that presents more than 24 hours after delivery requires a stepwise, multidisciplinary approach, Dr. Caroline Stella said at the annual meeting of the Society for Maternal-Fetal Medicine.

Postpartum headache affects 11%–80% of women after delivery, and usually is benign. Little has been known about headaches that start more than a day after delivery.

Dr. Stella of the University of Cincinnati and her associates retrospectively studied records for 95 women with severe headaches that started 25 hours to 32 days post partum and were unresponsive to usual doses of analgesics.

Approximately half of the women (47 patients) ultimately were diagnosed with tension-type or migraine headaches, and all responded to higher doses of analgesics or narcotics.

Twenty-three women (24%) with headaches caused by preeclampsia or eclampsia were treated with magnesium sulfate or antihypertensives; one had a cerebral venous thrombosis, and one had a subarachnoid hemorrhage.

Fifteen women (16%) had a final diagnosis of spinal headache. All received an anesthesiology consultation and initially were treated with analgesics. Twelve patients (13%) eventually required a blood patch. Headaches caused by other abnormalities in 10 women (11%) were treated with anticonvulsants, anticoagulants, or a dopamine antagonist.

Some patients had overlapping diagnoses. In 78 women, the headaches started while they were in the hospital; the other 17 were readmitted after postpartum discharge. Cesarean deliveries had been performed in 28 women.

Of the 95 women, 22 (23%) underwent cerebral imaging studies. Indications for imaging included persistent visual changes in 14 patients, focal neurologic deficits in 8, refractory headache in 8, and other reasons.

Neurologic symptoms other than headache resolved in 20 of the 22 women. Two patients were left with residual facial droop. No patients died.

On the basis of the study's findings, Dr. Stella and her associates recommended the following algorithm for work-up of postpartum headache presenting after 24 hours.

If neurologic deficits are present, perform cerebral imaging right away. If no neurologic deficits are present in a normotensive patient without proteinuria, treat for presumed migraine or tension-type headache.

If that therapy doesn't work and the patient has a history of regional anesthesia placement, consider the diagnosis of postdural puncture headache. Give intravenous fluids and analgesia, and consider administering a blood patch. If the headache still doesn't respond to therapy, perform cerebral imaging.

Presume a diagnosis of preeclampsia or eclampsia in women with hypertension or proteinuria and administer magnesium sulfate and antihypertensive agents. If the headache is unresponsive, perform cerebral imaging.

Eclampsia increased the risk for cognitive dysfunction years later in a separate case-control study of 87 women presented in a poster session by Annet M. Aukes, a medical student at the University of Groningen (the Netherlands).

From 6 to 8 years after their eclampsia episodes, 29 women were matched by age and year of pregnancy with formerly preeclamptic women and normotensive controls. All completed a validated Cognitive Failures Questionnaire designed to assess the likelihood of committing a cognitive error in an everyday task.

The women with a history of eclampsia scored significantly worse, compared with preeclamptic women or controls. The paradigm that eclamptic women can expect a full recovery should be reevaluated, she said in an interview.

“It's very important that these women are treated adequately right away when they come in with headaches so we can prevent more eclamptic seizures,” Ms. Aukes said.

These women need support months and years later to understand that their dysfunction is not psychological but caused by white matter lesions, she added.

MRIs of a patient's brain show cerebral edema (red arrows, left) in the occipital lobes and, on follow-up, complete resolution of the edema (red arrows, right). Photos courtesy Dr. Caroline Stella

SAN FRANCISCO — Evaluating persistent headache that presents more than 24 hours after delivery requires a stepwise, multidisciplinary approach, Dr. Caroline Stella said at the annual meeting of the Society for Maternal-Fetal Medicine.

Postpartum headache affects 11%–80% of women after delivery, and usually is benign. Little has been known about headaches that start more than a day after delivery.

Dr. Stella of the University of Cincinnati and her associates retrospectively studied records for 95 women with severe headaches that started 25 hours to 32 days post partum and were unresponsive to usual doses of analgesics.

Approximately half of the women (47 patients) ultimately were diagnosed with tension-type or migraine headaches, and all responded to higher doses of analgesics or narcotics.

Twenty-three women (24%) with headaches caused by preeclampsia or eclampsia were treated with magnesium sulfate or antihypertensives; one had a cerebral venous thrombosis, and one had a subarachnoid hemorrhage.

Fifteen women (16%) had a final diagnosis of spinal headache. All received an anesthesiology consultation and initially were treated with analgesics. Twelve patients (13%) eventually required a blood patch. Headaches caused by other abnormalities in 10 women (11%) were treated with anticonvulsants, anticoagulants, or a dopamine antagonist.

Some patients had overlapping diagnoses. In 78 women, the headaches started while they were in the hospital; the other 17 were readmitted after postpartum discharge. Cesarean deliveries had been performed in 28 women.

Of the 95 women, 22 (23%) underwent cerebral imaging studies. Indications for imaging included persistent visual changes in 14 patients, focal neurologic deficits in 8, refractory headache in 8, and other reasons.

Neurologic symptoms other than headache resolved in 20 of the 22 women. Two patients were left with residual facial droop. No patients died.

On the basis of the study's findings, Dr. Stella and her associates recommended the following algorithm for work-up of postpartum headache presenting after 24 hours.

If neurologic deficits are present, perform cerebral imaging right away. If no neurologic deficits are present in a normotensive patient without proteinuria, treat for presumed migraine or tension-type headache.

If that therapy doesn't work and the patient has a history of regional anesthesia placement, consider the diagnosis of postdural puncture headache. Give intravenous fluids and analgesia, and consider administering a blood patch. If the headache still doesn't respond to therapy, perform cerebral imaging.

Presume a diagnosis of preeclampsia or eclampsia in women with hypertension or proteinuria and administer magnesium sulfate and antihypertensive agents. If the headache is unresponsive, perform cerebral imaging.

Eclampsia increased the risk for cognitive dysfunction years later in a separate case-control study of 87 women presented in a poster session by Annet M. Aukes, a medical student at the University of Groningen (the Netherlands).

From 6 to 8 years after their eclampsia episodes, 29 women were matched by age and year of pregnancy with formerly preeclamptic women and normotensive controls. All completed a validated Cognitive Failures Questionnaire designed to assess the likelihood of committing a cognitive error in an everyday task.

The women with a history of eclampsia scored significantly worse, compared with preeclamptic women or controls. The paradigm that eclamptic women can expect a full recovery should be reevaluated, she said in an interview.

“It's very important that these women are treated adequately right away when they come in with headaches so we can prevent more eclamptic seizures,” Ms. Aukes said.

These women need support months and years later to understand that their dysfunction is not psychological but caused by white matter lesions, she added.

MRIs of a patient's brain show cerebral edema (red arrows, left) in the occipital lobes and, on follow-up, complete resolution of the edema (red arrows, right). Photos courtesy Dr. Caroline Stella

SAN FRANCISCO — Evaluating persistent headache that presents more than 24 hours after delivery requires a stepwise, multidisciplinary approach, Dr. Caroline Stella said at the annual meeting of the Society for Maternal-Fetal Medicine.

Postpartum headache affects 11%–80% of women after delivery, and usually is benign. Little has been known about headaches that start more than a day after delivery.

Dr. Stella of the University of Cincinnati and her associates retrospectively studied records for 95 women with severe headaches that started 25 hours to 32 days post partum and were unresponsive to usual doses of analgesics.

Approximately half of the women (47 patients) ultimately were diagnosed with tension-type or migraine headaches, and all responded to higher doses of analgesics or narcotics.

Twenty-three women (24%) with headaches caused by preeclampsia or eclampsia were treated with magnesium sulfate or antihypertensives; one had a cerebral venous thrombosis, and one had a subarachnoid hemorrhage.

Fifteen women (16%) had a final diagnosis of spinal headache. All received an anesthesiology consultation and initially were treated with analgesics. Twelve patients (13%) eventually required a blood patch. Headaches caused by other abnormalities in 10 women (11%) were treated with anticonvulsants, anticoagulants, or a dopamine antagonist.

Some patients had overlapping diagnoses. In 78 women, the headaches started while they were in the hospital; the other 17 were readmitted after postpartum discharge. Cesarean deliveries had been performed in 28 women.

Of the 95 women, 22 (23%) underwent cerebral imaging studies. Indications for imaging included persistent visual changes in 14 patients, focal neurologic deficits in 8, refractory headache in 8, and other reasons.

Neurologic symptoms other than headache resolved in 20 of the 22 women. Two patients were left with residual facial droop. No patients died.

On the basis of the study's findings, Dr. Stella and her associates recommended the following algorithm for work-up of postpartum headache presenting after 24 hours.

If neurologic deficits are present, perform cerebral imaging right away. If no neurologic deficits are present in a normotensive patient without proteinuria, treat for presumed migraine or tension-type headache.

If that therapy doesn't work and the patient has a history of regional anesthesia placement, consider the diagnosis of postdural puncture headache. Give intravenous fluids and analgesia, and consider administering a blood patch. If the headache still doesn't respond to therapy, perform cerebral imaging.

Presume a diagnosis of preeclampsia or eclampsia in women with hypertension or proteinuria and administer magnesium sulfate and antihypertensive agents. If the headache is unresponsive, perform cerebral imaging.

Eclampsia increased the risk for cognitive dysfunction years later in a separate case-control study of 87 women presented in a poster session by Annet M. Aukes, a medical student at the University of Groningen (the Netherlands).

From 6 to 8 years after their eclampsia episodes, 29 women were matched by age and year of pregnancy with formerly preeclamptic women and normotensive controls. All completed a validated Cognitive Failures Questionnaire designed to assess the likelihood of committing a cognitive error in an everyday task.

The women with a history of eclampsia scored significantly worse, compared with preeclamptic women or controls. The paradigm that eclamptic women can expect a full recovery should be reevaluated, she said in an interview.

“It's very important that these women are treated adequately right away when they come in with headaches so we can prevent more eclamptic seizures,” Ms. Aukes said.

These women need support months and years later to understand that their dysfunction is not psychological but caused by white matter lesions, she added.

MRIs of a patient's brain show cerebral edema (red arrows, left) in the occipital lobes and, on follow-up, complete resolution of the edema (red arrows, right). Photos courtesy Dr. Caroline Stella

SCOTTSDALE, ARIZ. — Think of coccidioidomycosis in patients with a rash, fever, and cough, even if they don't live in the southwestern United States where Coccidioides is endemic.

At least two patients have presented to the Mayo Clinic, Rochester, Minn., with cutaneous manifestations of coccidioidomycosis. Both patients were “snowbirds” who traveled to warmer climates in the southwest during the winter, according to physicians from the Mayo Clinic, Scottsdale, Ariz.

Although this mainly is a lung infection, cutaneous manifestations provide a clue to the diagnosis. “In the last 10 years at the Mayo Clinic in Arizona, I've been impressed by how often the dermatologist has a role to play in the diagnosis of coccidioidomycosis,” Dr. David J. DiCaudo said at a dermatology conference sponsored by Skin Disease Education Foundation.

The desert areas of the southwestern United States and northern Mexico are the prime locations of this fungus, which is found in the western United States, Central America, and south to Argentina. Most U.S. infections occur in Arizona and in California's San Joaquin Valley, where a syndrome with the infection was first recognized and dubbed “valley fever,” said Dr. DiCaudo of the Mayo Clinic, Scottsdale.

The incidence of coccidioidomycosis in Arizona more than tripled in the past decade, with a 56% increase in the past year alone. Droughts in recent years and construction activity stirring up soil and dust probably have contributed to the increase, he suggested. The organism lives in soil as filamentous mycelia that break down into arthroconidia, which can be carried on the wind and inhaled. Once inside people or animals, they transform into the spherule form recognized in biopsy specimens.

Most Coccidioides infections cause no symptoms. Around 40% of infected people develop a mild to moderate influenzalike illness with fever, cough, chills, and arthralgias. Even healthy people can be severely affected and laid low for weeks by the symptoms. Fewer than 1% develops severe infection or dissemination to the meninges or bones, with some deaths.

People of Filipino heritage are hundreds of times more likely to develop severe infection or dissemination, compared with the general population, and African Americans are at increased risk as well, Dr. DiCaudo said. People with compromised immune systems caused by pregnancy, HIV infection, or organ transplant, or those using steroids or other immunocompromising medications also face greater risk with this infection.

The painful red nodules of erythema nodosum are the most common cutaneous manifestation of coccidioidomycosis. They typically appear on the lower extremities 1–3 weeks after the onset of systemic symptoms and suggest a good prognosis.

Other cutaneous symptoms appear earlier. Acute exanthem may appear within the first 24–48 hours of illness. “I've seen several patients who had a florid eruption even before the onset of any other symptom. Days later, they developed fever and cough,” he said.

The acute exanthem can resemble a drug reaction. Associated pruritus may be mild to severe. Lesions on the palms are common. It may last days or weeks.

The infection also can cause Sweet's syndrome, presenting as painful plaques, often but not always on the upper body, associated with fever and peripheral blood leukocytosis. In other settings, Sweet's syndrome commonly is treated with systemic corticosteroids. “It's worth checking to make sure the patient doesn't have coccidioidomycosis first,” because an immunosuppressive drug would increase their risk, Dr. DiCaudo said.

Granulomatous dermatitis can develop early in the course of the disease with widely distributed papules and plaques.

All of these cutaneous symptoms are reactive conditions; no Coccidioides will be found in the skin. The skin symptoms evolve over a period of weeks or months as the patient recovers from the pulmonary infection.

A skin biopsy can be helpful, however, in rare disseminated infection, which typically develops 1–3 months after the onset of illness and can cause nodules, granulomatous plaques, and ulcers on the skin. It can mimic many other diseases including tuberculosis or acne. Even rarer is primary cutaneous infection at the site of inoculation, typically from injury by a laboratory pipette, a splinter, or even a cactus spine.

Serology is the key to diagnosing coccidioidomycosis. Keep in mind that the rash precedes seroconversion, so you may want to retest some patients with negative serologies 2 weeks later, he said. Low titers are common and shouldn't be dismissed.

The IgG antibody test can be positive and the IgM negative during active infection and shouldn't be interpreted as a past infection, he added. The antibodies tend to disappear following recovery, so a positive titer most likely represents acute infection.

The large spherules (10–80 mcm) are easily seen under microscopy, typically as granulomatous or suppurative inflammatory infiltrate. If needed, an in situ hybridization assay is available to distinguish the organism from Blastomyces or Cryptococcus.

Patients with coccidioidomycosis generally are managed by primary care physicians or infectious disease specialists.

Skin Disease Education Foundation and this news organization are wholly owned subsidiaries of Elsevier.

Sweet's syndrome, presenting as painful plaques, is associated with pulmonary coccidioidomycosis. Courtesy Dr. David J. DiCaudo

SCOTTSDALE, ARIZ. — Think of coccidioidomycosis in patients with a rash, fever, and cough, even if they don't live in the southwestern United States where Coccidioides is endemic.

At least two patients have presented to the Mayo Clinic, Rochester, Minn., with cutaneous manifestations of coccidioidomycosis. Both patients were “snowbirds” who traveled to warmer climates in the southwest during the winter, according to physicians from the Mayo Clinic, Scottsdale, Ariz.

Although this mainly is a lung infection, cutaneous manifestations provide a clue to the diagnosis. “In the last 10 years at the Mayo Clinic in Arizona, I've been impressed by how often the dermatologist has a role to play in the diagnosis of coccidioidomycosis,” Dr. David J. DiCaudo said at a dermatology conference sponsored by Skin Disease Education Foundation.

The desert areas of the southwestern United States and northern Mexico are the prime locations of this fungus, which is found in the western United States, Central America, and south to Argentina. Most U.S. infections occur in Arizona and in California's San Joaquin Valley, where a syndrome with the infection was first recognized and dubbed “valley fever,” said Dr. DiCaudo of the Mayo Clinic, Scottsdale.

The incidence of coccidioidomycosis in Arizona more than tripled in the past decade, with a 56% increase in the past year alone. Droughts in recent years and construction activity stirring up soil and dust probably have contributed to the increase, he suggested. The organism lives in soil as filamentous mycelia that break down into arthroconidia, which can be carried on the wind and inhaled. Once inside people or animals, they transform into the spherule form recognized in biopsy specimens.

Most Coccidioides infections cause no symptoms. Around 40% of infected people develop a mild to moderate influenzalike illness with fever, cough, chills, and arthralgias. Even healthy people can be severely affected and laid low for weeks by the symptoms. Fewer than 1% develops severe infection or dissemination to the meninges or bones, with some deaths.

People of Filipino heritage are hundreds of times more likely to develop severe infection or dissemination, compared with the general population, and African Americans are at increased risk as well, Dr. DiCaudo said. People with compromised immune systems caused by pregnancy, HIV infection, or organ transplant, or those using steroids or other immunocompromising medications also face greater risk with this infection.

The painful red nodules of erythema nodosum are the most common cutaneous manifestation of coccidioidomycosis. They typically appear on the lower extremities 1–3 weeks after the onset of systemic symptoms and suggest a good prognosis.

Other cutaneous symptoms appear earlier. Acute exanthem may appear within the first 24–48 hours of illness. “I've seen several patients who had a florid eruption even before the onset of any other symptom. Days later, they developed fever and cough,” he said.

The acute exanthem can resemble a drug reaction. Associated pruritus may be mild to severe. Lesions on the palms are common. It may last days or weeks.

The infection also can cause Sweet's syndrome, presenting as painful plaques, often but not always on the upper body, associated with fever and peripheral blood leukocytosis. In other settings, Sweet's syndrome commonly is treated with systemic corticosteroids. “It's worth checking to make sure the patient doesn't have coccidioidomycosis first,” because an immunosuppressive drug would increase their risk, Dr. DiCaudo said.

Granulomatous dermatitis can develop early in the course of the disease with widely distributed papules and plaques.

All of these cutaneous symptoms are reactive conditions; no Coccidioides will be found in the skin. The skin symptoms evolve over a period of weeks or months as the patient recovers from the pulmonary infection.

A skin biopsy can be helpful, however, in rare disseminated infection, which typically develops 1–3 months after the onset of illness and can cause nodules, granulomatous plaques, and ulcers on the skin. It can mimic many other diseases including tuberculosis or acne. Even rarer is primary cutaneous infection at the site of inoculation, typically from injury by a laboratory pipette, a splinter, or even a cactus spine.

Serology is the key to diagnosing coccidioidomycosis. Keep in mind that the rash precedes seroconversion, so you may want to retest some patients with negative serologies 2 weeks later, he said. Low titers are common and shouldn't be dismissed.

The IgG antibody test can be positive and the IgM negative during active infection and shouldn't be interpreted as a past infection, he added. The antibodies tend to disappear following recovery, so a positive titer most likely represents acute infection.

The large spherules (10–80 mcm) are easily seen under microscopy, typically as granulomatous or suppurative inflammatory infiltrate. If needed, an in situ hybridization assay is available to distinguish the organism from Blastomyces or Cryptococcus.

Patients with coccidioidomycosis generally are managed by primary care physicians or infectious disease specialists.

Skin Disease Education Foundation and this news organization are wholly owned subsidiaries of Elsevier.

Sweet's syndrome, presenting as painful plaques, is associated with pulmonary coccidioidomycosis. Courtesy Dr. David J. DiCaudo

SCOTTSDALE, ARIZ. — Think of coccidioidomycosis in patients with a rash, fever, and cough, even if they don't live in the southwestern United States where Coccidioides is endemic.

At least two patients have presented to the Mayo Clinic, Rochester, Minn., with cutaneous manifestations of coccidioidomycosis. Both patients were “snowbirds” who traveled to warmer climates in the southwest during the winter, according to physicians from the Mayo Clinic, Scottsdale, Ariz.

Although this mainly is a lung infection, cutaneous manifestations provide a clue to the diagnosis. “In the last 10 years at the Mayo Clinic in Arizona, I've been impressed by how often the dermatologist has a role to play in the diagnosis of coccidioidomycosis,” Dr. David J. DiCaudo said at a dermatology conference sponsored by Skin Disease Education Foundation.

The desert areas of the southwestern United States and northern Mexico are the prime locations of this fungus, which is found in the western United States, Central America, and south to Argentina. Most U.S. infections occur in Arizona and in California's San Joaquin Valley, where a syndrome with the infection was first recognized and dubbed “valley fever,” said Dr. DiCaudo of the Mayo Clinic, Scottsdale.

The incidence of coccidioidomycosis in Arizona more than tripled in the past decade, with a 56% increase in the past year alone. Droughts in recent years and construction activity stirring up soil and dust probably have contributed to the increase, he suggested. The organism lives in soil as filamentous mycelia that break down into arthroconidia, which can be carried on the wind and inhaled. Once inside people or animals, they transform into the spherule form recognized in biopsy specimens.

Most Coccidioides infections cause no symptoms. Around 40% of infected people develop a mild to moderate influenzalike illness with fever, cough, chills, and arthralgias. Even healthy people can be severely affected and laid low for weeks by the symptoms. Fewer than 1% develops severe infection or dissemination to the meninges or bones, with some deaths.

People of Filipino heritage are hundreds of times more likely to develop severe infection or dissemination, compared with the general population, and African Americans are at increased risk as well, Dr. DiCaudo said. People with compromised immune systems caused by pregnancy, HIV infection, or organ transplant, or those using steroids or other immunocompromising medications also face greater risk with this infection.

The painful red nodules of erythema nodosum are the most common cutaneous manifestation of coccidioidomycosis. They typically appear on the lower extremities 1–3 weeks after the onset of systemic symptoms and suggest a good prognosis.

Other cutaneous symptoms appear earlier. Acute exanthem may appear within the first 24–48 hours of illness. “I've seen several patients who had a florid eruption even before the onset of any other symptom. Days later, they developed fever and cough,” he said.

The acute exanthem can resemble a drug reaction. Associated pruritus may be mild to severe. Lesions on the palms are common. It may last days or weeks.

The infection also can cause Sweet's syndrome, presenting as painful plaques, often but not always on the upper body, associated with fever and peripheral blood leukocytosis. In other settings, Sweet's syndrome commonly is treated with systemic corticosteroids. “It's worth checking to make sure the patient doesn't have coccidioidomycosis first,” because an immunosuppressive drug would increase their risk, Dr. DiCaudo said.

Granulomatous dermatitis can develop early in the course of the disease with widely distributed papules and plaques.

All of these cutaneous symptoms are reactive conditions; no Coccidioides will be found in the skin. The skin symptoms evolve over a period of weeks or months as the patient recovers from the pulmonary infection.

A skin biopsy can be helpful, however, in rare disseminated infection, which typically develops 1–3 months after the onset of illness and can cause nodules, granulomatous plaques, and ulcers on the skin. It can mimic many other diseases including tuberculosis or acne. Even rarer is primary cutaneous infection at the site of inoculation, typically from injury by a laboratory pipette, a splinter, or even a cactus spine.

Serology is the key to diagnosing coccidioidomycosis. Keep in mind that the rash precedes seroconversion, so you may want to retest some patients with negative serologies 2 weeks later, he said. Low titers are common and shouldn't be dismissed.

The IgG antibody test can be positive and the IgM negative during active infection and shouldn't be interpreted as a past infection, he added. The antibodies tend to disappear following recovery, so a positive titer most likely represents acute infection.

The large spherules (10–80 mcm) are easily seen under microscopy, typically as granulomatous or suppurative inflammatory infiltrate. If needed, an in situ hybridization assay is available to distinguish the organism from Blastomyces or Cryptococcus.

Patients with coccidioidomycosis generally are managed by primary care physicians or infectious disease specialists.

Skin Disease Education Foundation and this news organization are wholly owned subsidiaries of Elsevier.

Sweet's syndrome, presenting as painful plaques, is associated with pulmonary coccidioidomycosis. Courtesy Dr. David J. DiCaudo

SAN FRANCISCO — Continued use of clopidogrel significantly decreased the risk of death or MI at 2 years in patients with intracoronary drug-eluting stents, compared with bare-metal stents or in patients with either kind of stent who stopped clopidogrel after 6 months of use, Dr. John S. MacGregor reported.

Patients with drug-eluting stents who stopped taking clopidogrel after 6 months of use fared the worst, compared with the other three groups, according to a study published online by investigators at Duke University, Durham, N.C., Dr. MacGregor said at a meeting sponsored by the California chapter of the American College of Cardiology.

All patients had been free of major cardiac events after stent implantation and 6 months of clopidogrel use. Patients with drug-eluting stents who stopped clopidogrel were more than twice as likely to die or have an MI by 24 months, compared with patients with drug-eluting stents who continued clopidogrel, said Dr. MacGregor of the University of California, San Francisco.

Rates of death or MI in patients with bare-metal stents who were on or off clopidogrel fell between rates for the drug-eluting-stent groups, and were significantly higher than in patients with drug-eluting stents who continued clopidogrel (JAMA 2007;297 [Epub doi:10.1001/jama.297.2.joc60179]).

Similar findings emerged for patients who were free of major cardiac events at 12 months after stent insertion. By 24 months, the rates of death or MI were 0% in patients with drug-eluting stents who continued clopidogrel, 5% in patients on drug-eluting stents who stopped clopidogrel or in patients with bare-metal stents who continued clopidogrel, and 4% in patients with bare-metal stents who stopped clopidogrel, he said at the meeting, also sponsored by the university.

Multiple investigators in the Duke study disclosed receiving research funding or having other financial ties with the companies that market clopidogrel, Sanofi-Aventis and Bristol-Myers Squibb.

Previous studies have shown that drug-eluting stents decrease the risk for major adverse cardiac events in the first 3–6 months after insertion, compared with bare-metal stents. Use of clopidogrel is recommended for 3 months after insertion of a sirolimus-coated stent or for 6 months after insertion of a paclitaxel-coated stent.

In a study presented at the 2006 American College of Cardiology meeting, however, the risk for death or MI, or for nonfatal MI, more than tripled in patients with drug-eluting stents who stopped taking clopidogrel, compared with patients with bare-metal stents who continued clopidogrel, he said. The 743 patients in the randomized Basal Stent Kosten Effektivitäts-Late Thrombotic Events (BASKET-LATE) trial had been event free after 6 months of clopidogrel use, and were followed for another 6–12 months off the drug.

These and other studies raise concerns that the recommended regimen of clopidogrel after insertion of drug-eluting stents is insufficient.

“My bias is to have people with drug-eluting stents take Plavix [clopidogrel] indefinitely unless they have bleeding problems,” Dr. MacGregor said.

Further studies are needed to determine the optimal duration of clopidogrel therapy after stent implantation and to include rates of side effects from extended-duration clopidogrel. Two out of three recent placebo-controlled trials of long-term clopidogrel use found that taking clopidogrel plus aspirin for 1–2.5 years significantly increased the rate of major or moderate bleeding events, compared with taking clopidogrel plus placebo, he said.

About 600,000 U.S. patients receive intracoronary stents each year. One year of clopidogrel therapy costs about $1,400, so therapy for 1 million patients would cost $1.4 billion per year, he noted. In addition, long-term clopidogrel use might interfere with valuable procedures such as colonoscopy or noncardiac surgery.

Patients with drug-eluting stents who stopped clopidogrel were more than twice as likely to die or have an MI. DR. MACGREGOR

SAN FRANCISCO — Continued use of clopidogrel significantly decreased the risk of death or MI at 2 years in patients with intracoronary drug-eluting stents, compared with bare-metal stents or in patients with either kind of stent who stopped clopidogrel after 6 months of use, Dr. John S. MacGregor reported.

Patients with drug-eluting stents who stopped taking clopidogrel after 6 months of use fared the worst, compared with the other three groups, according to a study published online by investigators at Duke University, Durham, N.C., Dr. MacGregor said at a meeting sponsored by the California chapter of the American College of Cardiology.

All patients had been free of major cardiac events after stent implantation and 6 months of clopidogrel use. Patients with drug-eluting stents who stopped clopidogrel were more than twice as likely to die or have an MI by 24 months, compared with patients with drug-eluting stents who continued clopidogrel, said Dr. MacGregor of the University of California, San Francisco.

Rates of death or MI in patients with bare-metal stents who were on or off clopidogrel fell between rates for the drug-eluting-stent groups, and were significantly higher than in patients with drug-eluting stents who continued clopidogrel (JAMA 2007;297 [Epub doi:10.1001/jama.297.2.joc60179]).

Similar findings emerged for patients who were free of major cardiac events at 12 months after stent insertion. By 24 months, the rates of death or MI were 0% in patients with drug-eluting stents who continued clopidogrel, 5% in patients on drug-eluting stents who stopped clopidogrel or in patients with bare-metal stents who continued clopidogrel, and 4% in patients with bare-metal stents who stopped clopidogrel, he said at the meeting, also sponsored by the university.

Multiple investigators in the Duke study disclosed receiving research funding or having other financial ties with the companies that market clopidogrel, Sanofi-Aventis and Bristol-Myers Squibb.

Previous studies have shown that drug-eluting stents decrease the risk for major adverse cardiac events in the first 3–6 months after insertion, compared with bare-metal stents. Use of clopidogrel is recommended for 3 months after insertion of a sirolimus-coated stent or for 6 months after insertion of a paclitaxel-coated stent.

In a study presented at the 2006 American College of Cardiology meeting, however, the risk for death or MI, or for nonfatal MI, more than tripled in patients with drug-eluting stents who stopped taking clopidogrel, compared with patients with bare-metal stents who continued clopidogrel, he said. The 743 patients in the randomized Basal Stent Kosten Effektivitäts-Late Thrombotic Events (BASKET-LATE) trial had been event free after 6 months of clopidogrel use, and were followed for another 6–12 months off the drug.

These and other studies raise concerns that the recommended regimen of clopidogrel after insertion of drug-eluting stents is insufficient.

“My bias is to have people with drug-eluting stents take Plavix [clopidogrel] indefinitely unless they have bleeding problems,” Dr. MacGregor said.

Further studies are needed to determine the optimal duration of clopidogrel therapy after stent implantation and to include rates of side effects from extended-duration clopidogrel. Two out of three recent placebo-controlled trials of long-term clopidogrel use found that taking clopidogrel plus aspirin for 1–2.5 years significantly increased the rate of major or moderate bleeding events, compared with taking clopidogrel plus placebo, he said.

About 600,000 U.S. patients receive intracoronary stents each year. One year of clopidogrel therapy costs about $1,400, so therapy for 1 million patients would cost $1.4 billion per year, he noted. In addition, long-term clopidogrel use might interfere with valuable procedures such as colonoscopy or noncardiac surgery.

Patients with drug-eluting stents who stopped clopidogrel were more than twice as likely to die or have an MI. DR. MACGREGOR

SAN FRANCISCO — Continued use of clopidogrel significantly decreased the risk of death or MI at 2 years in patients with intracoronary drug-eluting stents, compared with bare-metal stents or in patients with either kind of stent who stopped clopidogrel after 6 months of use, Dr. John S. MacGregor reported.

Patients with drug-eluting stents who stopped taking clopidogrel after 6 months of use fared the worst, compared with the other three groups, according to a study published online by investigators at Duke University, Durham, N.C., Dr. MacGregor said at a meeting sponsored by the California chapter of the American College of Cardiology.

All patients had been free of major cardiac events after stent implantation and 6 months of clopidogrel use. Patients with drug-eluting stents who stopped clopidogrel were more than twice as likely to die or have an MI by 24 months, compared with patients with drug-eluting stents who continued clopidogrel, said Dr. MacGregor of the University of California, San Francisco.

Rates of death or MI in patients with bare-metal stents who were on or off clopidogrel fell between rates for the drug-eluting-stent groups, and were significantly higher than in patients with drug-eluting stents who continued clopidogrel (JAMA 2007;297 [Epub doi:10.1001/jama.297.2.joc60179]).

Similar findings emerged for patients who were free of major cardiac events at 12 months after stent insertion. By 24 months, the rates of death or MI were 0% in patients with drug-eluting stents who continued clopidogrel, 5% in patients on drug-eluting stents who stopped clopidogrel or in patients with bare-metal stents who continued clopidogrel, and 4% in patients with bare-metal stents who stopped clopidogrel, he said at the meeting, also sponsored by the university.

Multiple investigators in the Duke study disclosed receiving research funding or having other financial ties with the companies that market clopidogrel, Sanofi-Aventis and Bristol-Myers Squibb.

Previous studies have shown that drug-eluting stents decrease the risk for major adverse cardiac events in the first 3–6 months after insertion, compared with bare-metal stents. Use of clopidogrel is recommended for 3 months after insertion of a sirolimus-coated stent or for 6 months after insertion of a paclitaxel-coated stent.

In a study presented at the 2006 American College of Cardiology meeting, however, the risk for death or MI, or for nonfatal MI, more than tripled in patients with drug-eluting stents who stopped taking clopidogrel, compared with patients with bare-metal stents who continued clopidogrel, he said. The 743 patients in the randomized Basal Stent Kosten Effektivitäts-Late Thrombotic Events (BASKET-LATE) trial had been event free after 6 months of clopidogrel use, and were followed for another 6–12 months off the drug.

These and other studies raise concerns that the recommended regimen of clopidogrel after insertion of drug-eluting stents is insufficient.

“My bias is to have people with drug-eluting stents take Plavix [clopidogrel] indefinitely unless they have bleeding problems,” Dr. MacGregor said.

Further studies are needed to determine the optimal duration of clopidogrel therapy after stent implantation and to include rates of side effects from extended-duration clopidogrel. Two out of three recent placebo-controlled trials of long-term clopidogrel use found that taking clopidogrel plus aspirin for 1–2.5 years significantly increased the rate of major or moderate bleeding events, compared with taking clopidogrel plus placebo, he said.

About 600,000 U.S. patients receive intracoronary stents each year. One year of clopidogrel therapy costs about $1,400, so therapy for 1 million patients would cost $1.4 billion per year, he noted. In addition, long-term clopidogrel use might interfere with valuable procedures such as colonoscopy or noncardiac surgery.

Patients with drug-eluting stents who stopped clopidogrel were more than twice as likely to die or have an MI. DR. MACGREGOR