Metformin for gestational diabetes in routine clinical practice.

AIMS To compare maternal and neonatal outcomes in women with gestational diabetes treated with diet, metformin and/or insulin in routine clinical practice in a single centre. METHODS We analysed prospectively collected data from the National Women's Health database for all women with gestational diabetes who delivered between January 2007 and December 2009. Since June 2007, women requiring medication have been given a choice of either metformin or insulin treatment, except women with a fetal abdominal circumference less than the 10th percentile, who were not offered metformin. RESULTS There were 1269 women with gestational diabetes; treatment was diet in 371, insulin in 399 and metformin in 465 (249 metformin alone, 216 metformin and insulin). Women treated with metformin and/or insulin had significantly higher BMIs compared with those in the diet group (P < 0.001) and had a higher fasting glucose at diagnosis (p < 0.001). Women treated with insulin had higher rates of Caesarean delivery (45.6% insulin, 37% metformin, 34% diet, P = 0.02) than women treated with metformin or diet. They also had higher rates of preterm births (19.2% insulin, 12.5% metformin, 12.1% diet, P = 0.005), customized large-for-gestational-age infants (18.5% insulin, 12.5% metformin, 12.4% diet, P = 0.02), neonatal admissions (18.7% insulin, 12.7% metformin, 14.0% diet, P = 0.04) and neonatal intravenous dextrose use (11.1% insulin, 5.1% metformin, 7.4% diet, P = 0.004). Neonatal outcomes were similar between diet- and metformin-treated women. CONCLUSIONS In routine practice, use of metformin in gestational diabetes was associated with fewer adverse outcomes compared with insulin, but baseline differences between treatment groups may have contributed to this.