|
|
| |||||||||||||||||||||||||||||||||||||||
Background Group B streptococci (Streptococcus agalactiae) are a major cause of meningitis and septicemia in neonates and pregnant women, but the importance of group B streptococcal disease in nonpregnant adults has not been clearly defined.
Methods We conducted a prospective surveillance of the pathogens responsible for meningitis for a period of 24 months in 35 hospitals and a referral laboratory in metropolitan Atlanta. We reviewed the clinical and laboratory records of all the nonpregnant adults identified as having invasive group B streptococcal disease during this period.
Results During 1989 and 1990 there were 424 patients with invasive group B streptococcal disease (annual incidence, 9.2 cases per 100,000 population). Of these patients, 46 percent were 1 month of age or younger, 6 percent were older than 1 month but younger than 18 years of age, and 48 percent were 18 or older. Men and nonpregnant women accounted for 68 percent (n = 140) of all cases among adults (annual incidence, 4.4 per 100,000). Clinical and laboratory records were available for 137.
In the nonpregnant adult patients (age, 18 to 99 years), the most common clinical diagnoses were skin, soft-tissue, or bone infection (in 36 percent); bacteremia with no identified source (30 percent); urosepsis (14 percent); pneumonia (9 percent); and peritonitis (7 percent). Risk factors included older age (
Conclusions Invasive group B streptococcal infection is a major problem not only in pregnant women and neonates but also in nonpregnant adults, especially those who are elderly and those who have chronic diseases.
60 years), the presence of diabetes mellitus, the presence of malignant neoplasms, and infection with the human immunodeficiency virus. The mortality rate in nonpregnant adults was 21 percent, accounting for 67 percent of all deaths related to group B streptococcal infection during the surveillance period.
Methods
Infections were diagnosed on the basis of the isolation of group B streptococci from normally sterile sites (e.g., blood, peritoneal fluid, and cerebrospinal fluid). Strains were identified as group B streptococci at local hospital laboratories with standard, commercially available diagnostic kits. All kits were based on the extraction of streptococcal-group antigens in soluble form and their identification with use of latex particles coated with group-specific antibodies. Patients with urinary tract infections and those with placental isolates were not included unless they had bacteremia. All the patients were living in the metropolitan Atlanta area at the time of infection. Surveillance was performed prospectively for 24 months (from January 1, 1989, to December 31, 1990) as part of an active bacteremia and meningitis surveillance project carried out in conjunction with the Georgia Department of Human Resources and the Centers for Disease Control and Prevention. The study area included all 35 hospitals and a major referral laboratory serving eight metropolitan Atlanta counties with a combined population of 2.3 million, including 1.65 million adults (those 18 years of age or older). Approximately 70 percent of the adults in the study area were white, 27 percent were black, and 3 percent were of other races or ethnic groups. During the study period, there were 131,534 pregnancies and 82,515 live births in the surveillance area (Vital Statistics, Georgia Department of Human Resources), and 6772 adults were estimated to be infected with the human immunodeficiency virus (HIV) (AIDS Registry, Georgia Department of Human Resources, Rimland D, and AIDS Research Consortium of Atlanta: unpublished data).
Initial case reports of invasive group B streptococcal disease were obtained from two independent sources: the hospital microbiology laboratories that provided isolates and hospital infection-control practitioners. Laboratory audits at all hospitals were performed every six months during the study period to evaluate reporting accuracy and to identify any cases that were not reported originally. The complete clinical and laboratory records of all men and nonpregnant women with invasive group B streptococcal disease were reviewed. Incidence calculations used population data obtained from the 1990 census. For estimations of relative risk, the age-specific prevalence of diabetes mellitus was obtained from the Georgia Department of Human Resources and the Diabetes Surveillance Report issued by the Centers for Disease Control and Prevention17. We calculated the prevalence of cancer in Atlanta by adjusting the incidence of cancer in metropolitan Atlanta in 1990 (figures provided by the American Cancer Society Tumor Registry) using a ratio of the national prevalence of cancer (reported by the National Cancer Institute)18 to the incidence of cancer in the United States in 1990 (figures provided by the American Cancer Society Tumor Registry). The prevalence was adjusted according to age groups with the use of a report of the Connecticut Tumor Registry19. Relative risk was calculated as previously described12.
Results
Invasive Group B Streptococcal Disease in the Metropolitan Atlanta Population
From January 1, 1989, through December 31, 1990, we identified 424 cases of invasive group B streptococcal disease in metropolitan Atlanta: 219 cases (52 percent) in children and 205 (48 percent) in adults. The annual incidence was 9.2 per 100,000 in the total population, 17 per 100,000 in children (including 2.6 cases in infants per 1000 live births), and 6.2 per 100,000 in adults. The annual incidence in nonpregnant adults was 4.4 per 100,000. Figure 1 shows the distribution of the infections according to age. The 140 nonpregnant adults who were infected were selected for further study.
|
The nonpregnant adults with group B streptococcal disease ranged in age from 18 to 99 years, with a mean age of 62 years; 22 (16 percent) of the infections occurred in patients 18 to 39 years of age (annual incidence, 1.1 per 100,000), 38 (27 percent) in patients 40 to 59 years of age (4.2 per 100,000), and 80 (57 percent) in patients 60 years of age or older (18 per 100,000). Women accounted for 32 percent of the cases among patients younger than 60 years of age, and for more than half the cases among the elderly ( 60 years of age). However, the overall rate of invasive group B streptococcal disease was slightly lower in women than in men (3.3 per 100,000 per year vs. 4.9 per 100,000 per year). The in-hospital mortality rate among nonpregnant adults was 21 percent (30 of a total of 45 deaths in all age groups [67 percent]).
The rate of invasive group B streptococcal infection was twice as high in black adults (6.5 per 100,000) as in white adults (3.2 per 100,000); the incidence increased with age and was particularly high in older blacks (Figure 2), although there was a relatively small black population older than 70 years (20,700). Seventy-three percent of the cases of group B streptococcal disease occurred in nonpregnant adult residents of two inner-city counties that accounted for only 52 percent of the total population. Adults residing in Fulton County, which includes the urban center of Atlanta, had a relative risk of group B streptococcal disease of 2.9, as compared with adults in the other seven counties (95 percent confidence interval, 2.1 to 4.0; P<0.001). In that county the risk was increased in both whites (relative risk, 2.6; 95 percent confidence interval, 1.6 to 4.1; P<0.001) and blacks (relative risk, 1.9; 95 percent confidence interval, 1.1 to 3.3; P 
0.04), as compared with whites and blacks in the other seven counties.
|
The clinical and laboratory records of 137 nonpregnant adults were available for review. Table 1 lists the clinical diagnoses. Skin, soft-tissue, and bone infections occurred in 50 (36 percent) of the patients and included cellulitis, foot ulcers, and decubitus ulcers. More than half the patients with such infections had diabetes mellitus (56 percent). Seven patients (all with diabetes mellitus) had osteomyelitis complicating a foot ulcer. Four patients with breast cancer had cellulitis of the arm or chest wall on the side on which a mastectomy had been performed 1 month to 10 years earlier. One of these patients had a breast implant infected with group B streptococcus removed. Four patients had scrotal cellulitis or abscess, two of whom were diabetics with penile implants (one implant was infected).
|
Group B streptococcus was cultured from the blood in 132 of the 140 nonpregnant adults (94 percent). Additional sites from which group B streptococcus was cultured include urine (18 patients), soft tissue or bone (12 patients), peritoneal fluid (7 patients), pleural fluid or sputum (4 patients), cerebrospinal fluid (4 patients), and synovial fluid (2 patients). In 42 patients, group B streptococcus was one of the agents in a polymicrobial bacteremia (Table 2). Staphylococci were the most common additional isolates from blood (29 cases) and were often the only other organism cultured (24 of 29 cases). The clinical diagnoses in patients with polymicrobial bacteremia were similar to those in patients with group B streptococcus as the only blood isolate.
|
Of the 137 patients for whom clinical records were available, all but 2 had one or more serious underlying diseases or conditions (Table 3), most commonly diabetes mellitus. Among the 43 diabetic patients, 19 required insulin and 16 did not require insulin; no information on insulin requirements was available for 8 patients. Neurologic disease resulting in alterations of mental status, paraplegia or quadriplegia, or other major deficits was present in 41 patients (30 percent). Acute or chronic renal failure (18 percent), liver disease (15 percent) (cirrhosis, hepatitis, and hepatic failure), and atherosclerotic vascular disease (cerebrovascular, coronary artery, and peripheral) were also common; 23 patients had malignant neoplasms (15 solid tumors and 8 hematologic cancers), including 5 patients with multiple myeloma. Eighteen patients (13 percent) had a history of urinary tract infections. Invasive group B streptococcal infection occurred in five HIV-infected patients and three additional patients with a history of intravenous drug use whose HIV status was unknown.
|
The well-defined study population allowed us to determine the relative risk of invasive group B streptococcal disease in specific groups. The age-specific incidence and relative risk of disease in adults with diabetes mellitus, cancer, and HIV infection are shown in Table 4. Persons infected with HIV and young adults with diabetes mellitus or cancer were at significantly increased risk of invasive group B streptococcal infections (28 to 30 times the risk in patients without such underlying conditions). The relative risk of infection in diabetics decreased with advancing age but remained significantly elevated in all age groups. The relative risk of group B streptococcal infection in patients with cancer also decreased with advancing age. However, in contrast to diabetics, elderly patients with cancer were at no greater risk than elderly patients without cancer.
|
We found group B streptococcal infection to be a significant and apparently increasing cause of invasive disease in nonpregnant adults. In our prospective, population-based study, infection in nonpregnant adults accounted for 66 percent of the total cases in adults. The annual incidence in adults of 4.4 per 100,000 is almost twice as high as that reported retrospectively in the same population six years earlier12. Importantly, 67 percent of all deaths related to group B streptococcal infection that occurred during the study occurred in men and nonpregnant women. Invasive group B streptococcal infections were seen with greater frequency in nonpregnant adults in our surveillance area than were cases of invasive disease due to Neisseria meningitidis, Listeria monocytogenes, or Haemophilus influenzae combined. The incidence approaches that of bacteremia due to Streptococcus pneumoniae in adults20.
In our large series, invasive group B streptococcal infections occurred almost exclusively in adults with serious underlying medical conditions. Although many of these conditions have previously been identified as risk factors,21,22,23 population-based data allow estimates of the risks for specific groups of adults. The risk was increased in older patients, patients with diabetes mellitus, and those with cancer. In addition, infection with HIV was associated with an age-specific risk of group B streptococcal infection that was 30 times the rate in the population without HIV infection. These data indicate that high-risk groups of nonpregnant adults as well as neonates and pregnant women should be a focus for improved prevention, diagnosis, and therapy of group B streptococcal disease.
The rate of invasive group B streptococcal infection was twice as high in black adults as in whites, and was particularly high in older blacks. Black race has been identified as an independent risk factor for both early-onset and late-onset group B streptococcal disease in neonates2. However, these associations may be due to socioeconomic factors. Most of the increase we observed in the black adults was associated with residence in Fulton County, the urban center of Atlanta. The evidence of a role for socioeconomic status is supported by the finding that whites living in the same area also had a significantly increased risk, as compared with whites living in the other counties studied. The increased prevalence of diabetes and other chronic conditions in blacks may also contribute to the increased rates of group B streptococcal disease.
Variable descriptions of the clinical spectrum and source of invasive group B streptococcal infection have resulted from hospital-based series that included only one or a few centers. These variations are most probably due to differences in patient populations. For example, the genitourinary tract was noted as a major source of group B streptococcal bacteremia by Bayer et al.,22 but not by Gallagher and Watanakunakorn13. Other series have emphasized pneumonia, endocarditis, polymicrobial bacteremia, nosocomial infection, and decubitus ulcers as primary clinical presentations in patients with invasive group B streptococcal disease13,14,15,21,22,23.
Population-based data provide a better perspective on the clinical spectrum of invasive group B streptococcal infection. In our study, skin and soft-tissue infections, including osteomyelitis, were the most common clinical presentation (36 percent). These infections were often complications of chronic diabetic or decubitus ulcers. Group B streptococcal bacteremia without any identified focus (30 percent) was the second most common presentation. One quarter of such patients with bacteremia had cirrhosis or hepatic failure, and 20 percent had renal failure. Many of the patients with group B streptococcal bacteremia of uncertain source also had indwelling intravenous catheters. In addition, polymicrobial bacteremia with group B streptococcus -- often in combination with Staphylococcus aureus -- was present in 41 percent, suggesting that catheter-related infection may be the focus in some cases.
Obstructive uropathy, recurrent urinary tract infections, and diabetes mellitus were the most common underlying illnesses in patients with group B streptococcal urosepsis. Peritonitis occurred as spontaneous bacterial peritonitis in patients with liver disease, as a complication of a ruptured abdominal viscus in some patients, or in patients undergoing long-term peritoneal dialysis. The majority of patients with pneumonia had documented aspiration or debilitating neurologic conditions that increase the likelihood of aspiration. The rate of nosocomial infections with group B streptococcus in our study (17 percent) was somewhat lower than the rates of 21 to 70 percent reported previously14,15,23.
Group B streptococci have been classified serologically into five major serotypes (Ia, Ia/c, Ib, II, and III) on the basis of differences in capsular polysaccharide. Work is under way to develop group B streptococcal vaccines4,7,8,9,10,11 directed against specific capsular polysaccharides associated with neonatal infection, particularly serotype III6,24. The group B streptococcal isolates in our study were not serotyped, and the limited data on the serotypes of isolates from adults are difficult to interpret. For example, of 25 group B streptococcal isolates from nonpregnant adults collected before and after the study from a major teaching hospital in our surveillance area, approximately 90 percent were Ia, Ia/c, or nontypable strains (Blumberg HM: personal communication). However, in a small sample from a multistate study,24 each of the serotypes was equally represented in adults with invasive disease, and in recent group B streptococcal isolates from other hospitals in our area, multiple serotypes were associated with disease in adults.
A major determinant of the susceptibility of infants to group B streptococcal disease is the level of capsular-type-specific antibodies in maternal serum7,25. Women who have group B streptococcal bacteremia post partum also have low levels of capsular-type-specific antibodies to the infecting strain25. However, the role of capsular-type-specific antibody levels in nonpregnant adults with invasive group B streptococcal disease remains unclear and must be evaluated before vaccines can be targeted to this population.
These population-based data indicate that invasive group B streptococcal infection is a major problem, not only in pregnant women and neonates but also in nonpregnant adults who are elderly or who have chronic diseases. Until now, the development of group B streptococcal vaccines has focused on perinatal disease. A clearer definition of the molecular epidemiology and host susceptibility in all risk groups is critical to future decisions about the development and use of these vaccines.
We are indebted to the infection-control practitioners, clinical microbiologists, and hospitals in Georgia Health District III for their help in identifying cases; to Dr. Henry M. Blumberg for helpful discussions and unpublished data; to Dr. David Rimland for data on HIV-infected patients; to Kathy Boring for helpful discussions and information on cancer prevalence; and to Dr. Anne LeMoine and Ms. Lane Pucko for assistance in the preparation of the manuscript.
Source Information
From the Department of Medicine, Emory University School of Medicine (M.M.F., R.C.H., T.S., D.S.S.); Veterans Affairs Medical Center (M.M.F., D.S.S.); Georgia Department of Human Resources (J.D.S.); and Meningitis and Special Pathogens Branch, Division of Bacterial and Mycotic Diseases, Centers for Disease Control and Prevention (A.S., J.D.W.) -- all in Atlanta.
Address reprint requests to Dr. Farley at the Veterans Affairs Medical Center (151), 1670 Clairmont Rd., Decatur, GA 30033.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Group B Streptococcal Disease in Adults
Mathew P., Justice A. C., Farley M. M., Stephens D. S., Wenger J. D., Wessels M. R., Kasper D. L.
Extract |
Full Text
N Engl J Med 1993;
329:1658-1659, Nov 25, 1993.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
