Treatment of Gestational Diabetes Mellitus
http://www.100md.com
《新英格兰医药杂志》
To the Editor: Crowther and colleagues ( June 16 issue)1 conclude that treatment of gestational diabetes mellitus reduces serious perinatal morbidity. However, we have some reservations.
For one, blood glucose monitoring and more attentive care of the subjects in the intervention group may have resulted in better outcomes, irrespective of insulin treatment. In addition, the incidence of postdate births was higher in the routine-care group; a postdate birth by itself carries a worse fetal prognosis than does delivery at 38 to 40 weeks of gestation.
Information on the degree of adherence of the intervention group to the study protocol (e.g., the percentage of the subjects who had reached the desired glucose levels) would provide critical support that the intervention itself caused the improved perinatal results.
Uriel Elchalal, M.D.
Amnon Brzezinski, M.D.
Hadassah Medical Center
91120 Jerusalem, Israel
elchalal@cc.huji.ac.il
References
Crowther CA, Hiller JE, Moss JR, McPhee AJ, Jeffries WS, Robinson JS. Effect of treatment of gestational diabetes mellitus on pregnancy outcomes. N Engl J Med 2005;352:2477-2486.
To the Editor: Crowther et al. report the results of their randomized trial designed to assess the effects of the treatment of gestational diabetes on the incidence of serious perinatal complications (defined as death, shoulder dystocia, bone fracture, and nerve palsy). They report a significant reduction in these events; the number needed to treat to prevent a serious outcome was 34 (absolute risk reduction, 3 percent). However, most of these events (64 percent) involved shoulder dystocia, which cannot be considered a serious complication: it is a subjective appreciation of a technical difficulty.
Second, in addition to the number needed to treat, one should consider the number needed to screen to avoid one event, in order to evaluate the effect of the systematic screening of all pregnant women in the general population. The price of this attitude (i.e., over-medicalization of a physiological condition) is detectable in the study itself: more induction of labor and more admissions to a neonatal nursery.
Thibault Richard, M.D.
H?pital Erasme
B-1070 Brussels, Belgium
thibrichard_truc@hotmail.com
Michel Vanhaeverbeek, M.D.
H?pital André Vésale
6110 Montigny-Le-Tilleul, Belgium
Amaryllis Haccuria, M.D.
H?pital Erasme
B-1070 Brussels, Belgium
The authors reply: Elchalal and Brzezinski raise the concern that factors other than glucose control may contribute to the better outcomes in the intervention group. Although this was not part of the protocol, women randomly assigned to the intervention group were more likely to have their labor induced, leading to an earlier gestational age at delivery. We interpret this as a "labeling effect," whereby physicians, concerned about potential complications, were more likely to induce labor in women known to have glucose intolerance.
Women in the intervention group received individualized dietary advice from a dietitian, education on self-monitoring their blood glucose levels, physician support, and ongoing obstetric care. Insulin therapy, according to the protocol, was introduced when glucose levels failed to reach desired levels. Blood glucose control was achieved with diet alone, without insulin therapy, in 80 percent of the women.
Richard and colleagues emphasize that 64 percent of the serious perinatal implications were from shoulder dystocia and note concern that this reflects a subjective judgment. For each woman enrolled in the trial, the primary caregiver completed a checklist that included the presence or absence of shoulder dystocia and the measures used (categorized according to severity) to overcome any dystocia. Mild measures (episiotomy or moderate traction) were used in 5 cases (22 percent), whereas in 11 cases (48 percent) the women required moderate traction (suprapubic pressure or McRobert's maneuver ), and in 7 cases (30 percent) more severe measures (rotation of the infant's shoulders or delivery of the posterior shoulder) were required.
The Australian Carbohydrate Intolerance Study in Pregnant Women (ACHOIS) trial was designed to assess whether treatment of gestational diabetes improved health outcomes; it was not designed as a screening trial. Populations vary as to the prevalence of pregnant women who will have gestational diabetes mellitus; hence the number needed to screen will need to be calculated for local populations.
Caroline A. Crowther, F.R.A.N.Z.C.O.G.
Janet E. Hiller, Ph.D.
Jeffrey S. Robinson, F.R.A.N.Z.C.O.G.
University of Adelaide
Adelaide, SA 5005, Australia
for the ACHOIS Trial Group
For one, blood glucose monitoring and more attentive care of the subjects in the intervention group may have resulted in better outcomes, irrespective of insulin treatment. In addition, the incidence of postdate births was higher in the routine-care group; a postdate birth by itself carries a worse fetal prognosis than does delivery at 38 to 40 weeks of gestation.
Information on the degree of adherence of the intervention group to the study protocol (e.g., the percentage of the subjects who had reached the desired glucose levels) would provide critical support that the intervention itself caused the improved perinatal results.
Uriel Elchalal, M.D.
Amnon Brzezinski, M.D.
Hadassah Medical Center
91120 Jerusalem, Israel
elchalal@cc.huji.ac.il
References
Crowther CA, Hiller JE, Moss JR, McPhee AJ, Jeffries WS, Robinson JS. Effect of treatment of gestational diabetes mellitus on pregnancy outcomes. N Engl J Med 2005;352:2477-2486.
To the Editor: Crowther et al. report the results of their randomized trial designed to assess the effects of the treatment of gestational diabetes on the incidence of serious perinatal complications (defined as death, shoulder dystocia, bone fracture, and nerve palsy). They report a significant reduction in these events; the number needed to treat to prevent a serious outcome was 34 (absolute risk reduction, 3 percent). However, most of these events (64 percent) involved shoulder dystocia, which cannot be considered a serious complication: it is a subjective appreciation of a technical difficulty.
Second, in addition to the number needed to treat, one should consider the number needed to screen to avoid one event, in order to evaluate the effect of the systematic screening of all pregnant women in the general population. The price of this attitude (i.e., over-medicalization of a physiological condition) is detectable in the study itself: more induction of labor and more admissions to a neonatal nursery.
Thibault Richard, M.D.
H?pital Erasme
B-1070 Brussels, Belgium
thibrichard_truc@hotmail.com
Michel Vanhaeverbeek, M.D.
H?pital André Vésale
6110 Montigny-Le-Tilleul, Belgium
Amaryllis Haccuria, M.D.
H?pital Erasme
B-1070 Brussels, Belgium
The authors reply: Elchalal and Brzezinski raise the concern that factors other than glucose control may contribute to the better outcomes in the intervention group. Although this was not part of the protocol, women randomly assigned to the intervention group were more likely to have their labor induced, leading to an earlier gestational age at delivery. We interpret this as a "labeling effect," whereby physicians, concerned about potential complications, were more likely to induce labor in women known to have glucose intolerance.
Women in the intervention group received individualized dietary advice from a dietitian, education on self-monitoring their blood glucose levels, physician support, and ongoing obstetric care. Insulin therapy, according to the protocol, was introduced when glucose levels failed to reach desired levels. Blood glucose control was achieved with diet alone, without insulin therapy, in 80 percent of the women.
Richard and colleagues emphasize that 64 percent of the serious perinatal implications were from shoulder dystocia and note concern that this reflects a subjective judgment. For each woman enrolled in the trial, the primary caregiver completed a checklist that included the presence or absence of shoulder dystocia and the measures used (categorized according to severity) to overcome any dystocia. Mild measures (episiotomy or moderate traction) were used in 5 cases (22 percent), whereas in 11 cases (48 percent) the women required moderate traction (suprapubic pressure or McRobert's maneuver ), and in 7 cases (30 percent) more severe measures (rotation of the infant's shoulders or delivery of the posterior shoulder) were required.
The Australian Carbohydrate Intolerance Study in Pregnant Women (ACHOIS) trial was designed to assess whether treatment of gestational diabetes improved health outcomes; it was not designed as a screening trial. Populations vary as to the prevalence of pregnant women who will have gestational diabetes mellitus; hence the number needed to screen will need to be calculated for local populations.
Caroline A. Crowther, F.R.A.N.Z.C.O.G.
Janet E. Hiller, Ph.D.
Jeffrey S. Robinson, F.R.A.N.Z.C.O.G.
University of Adelaide
Adelaide, SA 5005, Australia
for the ACHOIS Trial Group