Does intrapartum antibiotic prophylaxis decrease the incidence of maternal group B streptococcal infections?

Effect of intrapartum antibiotic prophylaxis against group B streptococcal infection on comparisons of rates of endometritis and urinary tract infection in multicenter surveillance

OBJECTIVE

To establish whether antibiotic prophylaxis against group B streptococcal infection may be a confounding factor in comparisons of rates of endometritis and urinary tract infection after vaginal delivery.

DESIGN

Prospective study.

SETTING

Maternity units at 48 hospitals in a regional surveillance network in France during 2001-2004.

METHODS

The maternity units used a common protocol to establish whether antibiotic prophylaxis was indicated. Risk factors for endometritis and urinary tract infections were evaluated using multiple logistic regression.

RESULTS

We analyzed 49,786 vaginal deliveries. The percentage of women receiving antibiotic prophylaxis varied widely and significantly among the maternity units (range, 4.4%-26.0%; median, 15.8%; 25th percentile, 12.1%; 75th percentile, 19.0%) (P < .001, by Mantel-Haenszel chi(2) test). The incidence rate of endometritis was significantly reduced from 0.25% to 0.11% by antibiotic prophylaxis (P = .001). There was a decrease in the incidence of urinary tract infection from 0.37% to 0.32%, but it was not statistically significant (P = .251).

CONCLUSIONS

A reduction in the incidence of endometritis was observed when intrapartum antibiotic prophylaxis against group B streptococcal infection was used. However, the proportion of women considered to be at risk of infection varied widely among institutions. Comparisons of rates of endometritis among maternity units, but not urinary tract infection rates, should take into account antibiotic prophylaxis as a significant confounding factor.

Decreased rates of nosocomial endometritis and urinary tract infection after vaginal delivery in a French surveillance network, 1997-2003

OBJECTIVES

To identify independent risk factors for endometritis and urinary tract infection (UTI) after vaginal delivery, and to monitor changes in nosocomial infection rates and derive benchmarks for prevention.

DESIGN

Prospective study.

METHODS

We analyzed routine surveillance data for all vaginal deliveries between January 1997 and December 2003 at 66 maternity units participating in the Mater Sud-Est surveillance network. Adjusted odds ratios for risk of endometritis or UTI were obtained using a logistic regression model.

RESULTS

The overall incidence rates were 0.5% for endometritis and 0.3% for UTI. There was a significant decrease in the incidence and risk of endometritis but not of UTI during the 7-year period. Significant risk factors for endometritis were fever during labor, parity of 1, and instrumental delivery and/or manual removal of the placenta. Significant risk factors for UTI were urinary infection on admission, premature rupture of membranes (more than 12 hours before admission), blood loss of more than 800 mL, parity of 1, instrumental delivery, and receipt of more than 5 vaginal digital examinations. Each maternity unit received a poster showing graphs of the number of expected and observed cases of UTI and endometritis associated with vaginal deliveries, which enabled each maternity unit to determine their rank within the network and to initiate prevention programs.

CONCLUSIONS

Although routine surveillance means additional work for maternity units, our results demonstrate the usefulness of regular targeted monitoring of risk factors and of the most common nosocomial infections in obstetrics. Most of the information needed for monitoring is already present in the patients' records.