Intrapartum Antibiotic Use Linked to E coli Infection?
Intrapartum Antibiotic Use Linked to E coli Infection?
Schrag SJ, Hadler JL, Arnold KE, Martell-Cleary P, Reingold A, Schuchat A
Pediatrics. 2006;118(2):570-576
The authors note that the use of intrapartum prophylactic antibiotics in mothers at high risk for group B streptococcal (GBS) infections has markedly reduced the rate of invasive GBS infection in newborns. One concern has been that Escherichia coli (E coli) might expand to fill the "ecological niche" occurring with reduced GBS infections. Data have been mixed as to whether there has been an increase in E coli infection or E coli resistance to ampicillin, another possible effect of using that drug for intrapartum prophylaxis.
This report utilized 2 study methods. The first was an evaluation of data in the Centers for Disease Control and Prevention (CDC) Active Bacterial Core Surveillance Emerging Infections Program Network (a multistate surveillance program) looking at rates of invasive bacterial infections. The second part of the report included case-control analysis of patients with invasive E coli infection to determine risk factors associated with that infection. The authors focused on early-onset E coli infections (occurring ≤ 6 days of life) of the spinal fluid or blood.
Data were collected during the period from 1997-2001. Over 1000 infants born in the surveillance hospitals served as controls. The authors included the following risk factors in their analyses: maternal race and ethnicity, delivery method, gestational age of the infant, duration of membrane rupture, the presence of fever during the intrapartum period, and whether the mother received intrapartum antibiotics.
During the study period, there were 132 cases of early-onset E coli infection, with 61% occurring on the day of birth. Sixteen percent of the infants died. Fifty-three percent of the 132 cases had been exposed to intrapartum antibiotics. The cases were 58% male and 52% white. One third of the cases weighed < 1500 grams, but 40% weighed at least 2500 grams. Almost 50% of cases were ≤ 33 weeks gestational age. Ninety-five percent of the cases had culture-proven bacteremia, and 5% had culture-proven meningitis. Just over one third (36%) of the cases had ampicillin-sensitive isolates, while 52% had ampicillin-resistant isolates (12% had unknown sensitivity results). This means that 59% of those who had sensitivity testing were infected with ampicillin-resistant isolates.
In the analysis of the 68 ampicillin-resistant vs 48 ampicillin-sensitive cases, 2 factors were associated with ampicillin-resistant E coli. Infants born after membrane rupture of 18 hours or longer (adjusted OR 2.5, 95% CI 1.1-5.7) or whose mothers had received 2 or more doses of intrapartum penicillin or ampicillin prophylaxis (adjusted OR 5.3, 95% CI 1.6-17.2) had an increased risk of ampicillin-resistant E coli infection. Simple receipt of intrapartum antibiotics (all infants who were exposed, not just those with several doses) was not associated with resistant E coli infection.
In case-control analysis, the following factors were associated with early-onset E coli infection: gestational age ≤ 33 weeks (adjusted OR 26.5), the presence of intrapartum fever (adjusted OR 6.6), and membrane rupture of ≥ 18 hours' duration (adjusted OR 3.5). In the case-control analyses, neither the duration of antibiotic exposure nor the number of doses of penicillin or ampicillin was associated with having invasive E coli. The risk factors for having ampicillin-resistant invasive E coli were the same, and the odds ratios were of similar magnitude.
The authors make a very valid point: the analysis of only cases (divided into those with sensitive vs resistant E coli isolates) raised the concern of antibiotic use leading to E coli infections, but such an analytic approach does not include any potential cases that were prevented by prophylaxis. The case-control analysis compensates by comparing resistant infections to noninfections, and the authors found that only the nonprophylaxis variables were associated with either infection or noninfection. The authors point out that, for now, their data do not suggest that intrapartum antibiotic prophylaxis raises the risk of E coli or resistant E coli infection. However, the evolving nature of bacterial resistance means that such analyses will need to be repeated periodically, so stay tuned.
Abstract
Schrag SJ, Hadler JL, Arnold KE, Martell-Cleary P, Reingold A, Schuchat A
Pediatrics. 2006;118(2):570-576
The authors note that the use of intrapartum prophylactic antibiotics in mothers at high risk for group B streptococcal (GBS) infections has markedly reduced the rate of invasive GBS infection in newborns. One concern has been that Escherichia coli (E coli) might expand to fill the "ecological niche" occurring with reduced GBS infections. Data have been mixed as to whether there has been an increase in E coli infection or E coli resistance to ampicillin, another possible effect of using that drug for intrapartum prophylaxis.
This report utilized 2 study methods. The first was an evaluation of data in the Centers for Disease Control and Prevention (CDC) Active Bacterial Core Surveillance Emerging Infections Program Network (a multistate surveillance program) looking at rates of invasive bacterial infections. The second part of the report included case-control analysis of patients with invasive E coli infection to determine risk factors associated with that infection. The authors focused on early-onset E coli infections (occurring ≤ 6 days of life) of the spinal fluid or blood.
Data were collected during the period from 1997-2001. Over 1000 infants born in the surveillance hospitals served as controls. The authors included the following risk factors in their analyses: maternal race and ethnicity, delivery method, gestational age of the infant, duration of membrane rupture, the presence of fever during the intrapartum period, and whether the mother received intrapartum antibiotics.
During the study period, there were 132 cases of early-onset E coli infection, with 61% occurring on the day of birth. Sixteen percent of the infants died. Fifty-three percent of the 132 cases had been exposed to intrapartum antibiotics. The cases were 58% male and 52% white. One third of the cases weighed < 1500 grams, but 40% weighed at least 2500 grams. Almost 50% of cases were ≤ 33 weeks gestational age. Ninety-five percent of the cases had culture-proven bacteremia, and 5% had culture-proven meningitis. Just over one third (36%) of the cases had ampicillin-sensitive isolates, while 52% had ampicillin-resistant isolates (12% had unknown sensitivity results). This means that 59% of those who had sensitivity testing were infected with ampicillin-resistant isolates.
In the analysis of the 68 ampicillin-resistant vs 48 ampicillin-sensitive cases, 2 factors were associated with ampicillin-resistant E coli. Infants born after membrane rupture of 18 hours or longer (adjusted OR 2.5, 95% CI 1.1-5.7) or whose mothers had received 2 or more doses of intrapartum penicillin or ampicillin prophylaxis (adjusted OR 5.3, 95% CI 1.6-17.2) had an increased risk of ampicillin-resistant E coli infection. Simple receipt of intrapartum antibiotics (all infants who were exposed, not just those with several doses) was not associated with resistant E coli infection.
In case-control analysis, the following factors were associated with early-onset E coli infection: gestational age ≤ 33 weeks (adjusted OR 26.5), the presence of intrapartum fever (adjusted OR 6.6), and membrane rupture of ≥ 18 hours' duration (adjusted OR 3.5). In the case-control analyses, neither the duration of antibiotic exposure nor the number of doses of penicillin or ampicillin was associated with having invasive E coli. The risk factors for having ampicillin-resistant invasive E coli were the same, and the odds ratios were of similar magnitude.
The authors make a very valid point: the analysis of only cases (divided into those with sensitive vs resistant E coli isolates) raised the concern of antibiotic use leading to E coli infections, but such an analytic approach does not include any potential cases that were prevented by prophylaxis. The case-control analysis compensates by comparing resistant infections to noninfections, and the authors found that only the nonprophylaxis variables were associated with either infection or noninfection. The authors point out that, for now, their data do not suggest that intrapartum antibiotic prophylaxis raises the risk of E coli or resistant E coli infection. However, the evolving nature of bacterial resistance means that such analyses will need to be repeated periodically, so stay tuned.
Abstract
Source...