Pregnant women are at risk of infection during labor and delivery; most infections of the female pelvic organs occur when normal flora of the female genital or gastrointestinal tract contaminate the normally sterile amniotic fluid and uterus. These bacteria can become pathogenic, particularly in devitalized tissue [1]. Bacterial infections in pregnant women may also originate from inoculation of bacteria during childbirth or surgery or from hematogenous spread [2]. Uterine infections are much more common after Cesarean delivery (C-section) than following a vaginal delivery [3]. Fortunately, postpartum or puerperal infections have decreased over the last hundred years due to improved procedures and effective antibiotic use. However, infections still cause about 13% of pregnancy-related deaths and are the fifth leading cause of death among this population [1].

Group A Streptococcus (GAS), or Streptococcus pyogenes, is an uncommon, but serious and potentially preventable cause of postpartum and postsurgical infections. The incidence of confirmed GAS infections among deliveries in a study in Jerusalem over 6.5 years ranged from 0.33 to 3.16 per 1000 births; suspected intra- or postpartum GAS infections occurred in 13.2% of all deliveries [4]. In a review of records of all mothers and babies with GAS bacteremia in the United Kingdom from 1980–1999, peripartum GAS infection occurred in 1 of every 11,000 live births [5]. The source of a sporadic postpartum GAS infection is typically unknown, but outbreaks of postpartum and postsurgical GAS infections have been associated with colonized healthcare workers. Healthcare workers who were asymptomatic carriers of GAS have been identified in 15 of 21 outbreaks of postpartum and postsurgical infections reported from 1976 through 2005 [6–26]; these infections represent a preventable manifestation of severe GAS disease.

In 2002, Centers for Disease Control (CDC) published guidelines for prevention of secondary invasive GAS infections, including those among postpartum and postsurgical patients [27]. Recommended strategies for the investigation of both a single case and clusters of postpartum or postsurgical GAS infection were developed. In brief, a single healthcare-associated infection should trigger notification of the hospital infection control practitioner, enhanced surveillance for additional cases and storage of GAS patient isolates for potential strain characterization. When two or more postpartum or postsurgical cases occur in the same healthcare facility within six months, CDC recommends that epidemiologic links between the cases be investigated, including screening for GAS colonization of healthcare personnel linked to known cases.

Although these guidelines were published in 2002, the level of awareness and adherence to these recommendations is unknown. Also, clinician knowledge and management of pregnancy-related and postsurgical infections has not been assessed. We conducted a survey of obstetric and gynecologic providers (OB/GYNs) to estimate the frequency of intrapartum, postpartum, and postsurgical infections they encounter to determine bacterial culturing practices among providers, and to estimate the frequency of GAS isolation among the respondents’ patient populations. We also sought to estimate awareness and adherence to the CDC guidelines, assess barriers to follow the guidelines, and query providers regarding more effective methods for distributing these recommendations.

In July 2005, an anonymous questionnaire focusing on current practices regarding intrapartum, postpartum, and postsurgical infections was mailed to 1300 American College of Obstetricians and Gynecologists (ACOG) Fellows and Junior Fellows in Practice; 1000 subjects were members of the Collaborative Ambulatory Research Network (CARN). Members of CARN are practicing OB/GYNs who have volunteered to participate regularly in surveys, facilitating assessment of clinical practice patterns, and aiding development of educational materials. The remaining 300 subjects consisted of a computer-generated random sample of ACOG Fellows and Junior Fellows in practice who are practicing obstetrics and/or gynecology and had not received a survey from ACOG during the previous two years (non-CARN). A second mailing was sent approximately five weeks after the first to those individuals who did not initially respond. A final reminder mailing was sent approximately six weeks later. All survey respondents who reported performing deliveries or surgeries in 2004 were eligible for inclusion in our analysis.

The survey included questions about the practitioner's age and training, his/her practice, patient population demographics, and specific questions assessing experience with diagnosing and treating intrapartum, postpartum (e.g., wound infections, endometritis, bacteremia), and postsurgical infections (e.g., wound infections, endometritis, vaginal cuff cellulitis, bacteremia), particularly infections potentially due to GAS. Survey respondents were asked to estimate the racial and ethnic distributions of their patient population, number of deliveries and surgeries performed and related infections seen during 2004. Questions also assessed the practitioner's use and understanding of the CDC guidelines for prevention of GAS infections among postpartum and postsurgical patients. Question formats were primarily multiple choice or opinion questions using a scaled response.

Survey responses were double-entered into an American Standard Code for Information Interchange (ASCII) file, corrected for data entry errors, and converted into an SAS 9 database (SAS Institute, Cary, NC, USA). The racial and ethnic characteristics of the provider's patient population were dichotomized into the following categories, based on the distribution of responses: less than 80% white versus 80% or more white; less than 25% versus 25% or more black; less than 20% versus 20% or more Hispanic; less than 7% versus 7% or more Asian/Pacific Islander. The percentage of patients with Medicaid was divided into low (<25%), medium (25–49%), and high (≥50%) categories. Differences in categorical measures were assessed using the chi-square test. The Wilcoxon 2-sample test was used to compare medians among continuous measures. All analyses were tested for significance using a P value of <.05. Comparisons of responses by CARN and non-CARN members were completed for all survey questions. All statistically significant differences between CARN and non-CARN are presented; otherwise the results for both CARN and non-CARN combined are shown.

Our study is the first published survey of the knowledge, attitudes, and practices of OB/GYNs regarding the perceived frequency, etiology, and management of intrapartum, postpartum, and postsurgical infections. Among a large number of OB/GYNs, we found that most empirically treat a variety of infections encountered in their routine practice. Although this approach is within current standards of care, this presents a missed opportunity for the identification and potential prevention of illnesses such as severe GAS infections.

Both our survey and the current literature suggest that intraamniotic, postpartum, and postsurgical infections are not uncommon. The providers in our survey identify approximately three postpartum infections per 100 vaginal deliveries and 7 postsurgical infections per 100 Cesarean deliveries and hysterectomies. These estimates are conservative, as they do not include deliveries by other providers (e.g., family practitioners), but are consistent with previous estimates in the literature. Intraamniotic infections have been reported to range from 0.5% to 10.5% of all pregnancies, occurring in 1–5% of term pregnancies and up to 10% in patients with premature labor [28–32]. Puerperal or postpartum endometritis has been reported to occur in 1–3% of women with vaginal deliveries and among 5–15% of women having scheduled repeat Cesarean deliveries [1, 28].

Nearly all survey respondents treat intrapartum, postpartum, and postsurgical infections empirically although providers are more likely to obtain diagnostic specimens from patients with postsurgical infections. While no formalized guidelines on diagnostic management of intrapartum and postpartum infections exist, empiric therapy is not contrary to current practice recommendations. Recently developed ACOG guidelines on the use of prophylactic antibiotics in labor and delivery recommend the use of prophylactic antibiotics in both high- and low-risk patients undergoing Cesarean delivery although the data to support use among low-risk patients is inconclusive [33]. Blood cultures are typically taken only when the patient does not respond to empiric antibiotics or if complications arise [34]. The rationale for not obtaining diagnostic cultures is multifold; antibiotic therapy is typically empiric and based on clinical diagnosis, patients often respond to antibiotics before culture results are known, anaerobes are notoriously difficult to isolate, accurate postpartum endometrial specimens are difficult to obtain because of possible contamination by the lower genital tract, and cultures add cost and time [35].

Conversely, pretreatment cultures facilitate management of patients who fail initial empiric antibiotic [36]. Tailoring antimicrobial therapy is impossible without identifying the etiologic agent. Investigations of the specific pathogens causing intraamniotic, postpartum, and postsurgical infections have shown that the primary causative agents change over time. For example, the principal etiology of early onset neonatal sepsis has ranged from GAS in the 1930s–1940s, E. coli in the 1940s–1970s, to group B streptococci in the current era [37]. Also, recent studies have documented an increase in the proportion of healthcare-associated bloodstream infections and the emergence of community-associated methicillin-resistant Staphylococcus aureus infections among pregnant and postsurgical patients [38–40]. Comprehensive studies to evaluate the etiologies of pregnancy-related infections have not been performed for over 15 years [41–45]. Periodic evaluations of the etiologies of pregnancy-related infections and the prevalence of antibiotic-resistant isolates among this population would be useful.

Another benefit of obtaining pretreatment cultures is that identification of certain pathogens, such as GAS, should trigger specific infection control measures to prevent spread of disease. GAS infections can be devastating with an overall case fatality ratio (CFR) of 13% and CFRs of 25–36% for the most severe manifestations—necrotizing fasciitis and streptococcal toxic shock syndrome [46]. These infections are not rare. Because we only surveyed ~3% of the nation's OB/GYNs, we estimate that as many as 2700 postpartum and 2700 postsurgical GAS infections may occur annually in the United States. Healthcare-associated transmission of GAS infections can sometimes be prevented, for example, in instances where an asymptomatic colonized healthcare worker is identified and treated as a result of a thorough epidemiologic investigation [27].

Provider responses to the GAS-specific questions included in our survey underscore both the lack of awareness of the CDC GAS guidelines among one of their targeted audiences—OB/GYNs—and potential means to correct this information gap. Very few providers were aware of the guidelines and the recommended public health action following identification of postpartum and postsurgical GAS infections. The principal reason for not understanding or following these guidelines was a lack of awareness of their existence, most likely because they were published in a journal not read by this group of providers. Potential solutions to this dilemma include publication in journals relevant to OB-GYNs, incorporation into ACOG guidelines and hospital infection control protocols, presentation at relevant meetings, and inclusion in CME lecture series.

A common thread in our survey results is the association of increasing years of experience and older age with a variety of practice patterns. Older, more experienced providers were far more likely to attempt to determine the etiology of intra- and postpartum infections and also had different beliefs from younger providers as to the specific etiologies of such infections.The reasons for these trends are unknown but may include changes in medical education or in the epidemiology of these infections.

Limitations of our survey include our low response rate, particularly among non-CARN providers although it is consistent with other ACOG surveys [47, 48]. An important limitation is that we were unable to validate the estimated number of procedures performed and actual practices reported by the respondents. It is possible that the frequencies of infections reported are significantly underestimated. Also, given the low number of providers who were aware of the CDC guidelines for postpartum and postsurgical GAS infections, attempts to estimate the respondents' understanding of specific aspects of these guidelines are of limited value.

In summary, although the use of empiric antibiotics in intraamniotic, postpartum, and postsurgical infections may currently be effective in most cases, the paucity of diagnostic cultures obtained presents a missed opportunity to monitor the etiology of these infections and to identify a potentially preventable cause of serious healthcare-associated disease—GAS infections. Current efforts to educate providers regarding the settings in which the identification of GAS should trigger a public health response and augment infection control practices are inadequate. However, our survey identified more effective means of communicating with practicing OB/GYNs and also emphasized that educational efforts should target the infection control practitioner (ICP) as most hospitals currently rely on the ICP to identify potential healthcare-associated infections and initiate necessary investigations. Periodic, time-limited studies of the etiologies of pregnancy-related infections by public health and clinical researchers would help to monitor changes in the principal pathogens, track trends in antimicrobial resistance, and guide clinical management.