The Minnesota Department of Health conducts statewide, population-based surveillance for GBS disease as part of the Centers for Disease Control and Prevention Active Bacterial Core Surveillance Network/Emerging Infections Program. Invasive disease is defined as isolation of GBS bacteria from a normally sterile site, such as blood, pleural fluid, cerebrospinal fluid, joint fluid, or bone (3
). To ensure completeness of reporting, the Minnesota Department of Health audits laboratories to identify all GBS bacteria–positive cultures from normally sterile sites. For each case, a standardized case report form is completed by hospital infection control practitioners. GBS isolates are sent to the Minnesota Department of Health Public Health Laboratory for susceptibility testing using broth microdilution. Erythromycin-resistant, clindamycin-susceptible isolates are tested for inducible clindamycin resistance by double-disk diffusion (D test). Interpretation is based on Clinical and Laboratory and Standard Institute protocols (4
). Serotyping is performed at the Centers for Disease Control and Prevention by latex agglutination tests with rabbit antiserum to GBS capsular polysaccharide types Ia, Ib, and II–VIII (5
). When latex tests are indeterminate, the Lancefield method is used (6
The study comprised all Minnesota residents aged >65 years with invasive GBS disease during 2003–2007. LTCF residence was defined as living in an LTCF before the date of first positive culture. Resident addresses were checked by a reverse-address directory to determine whether they corresponded with the address of an LTCF. All other residents were defined as community dwelling. Incidence was calculated using 2000 census data. Analyses were conducted using SAS version 9.1 (SAS Institute, Cary, NC, USA); the χ2 test was used to evaluate differences in proportions for discrete variables.
A total of 723 cases of invasive GBS disease among persons >65 years of age were reported; 596 (82.4%) cases occurred among community residents, and 127 (17.6%) occurred among LTCF residents (). The overall incidence rate was 24.3 cases per 100,000 persons. Incidence did not vary significantly by year but did increase with age (19.3/100,000 at 65–74 years, 26.3/100,000 at 75–84 years, and 36.9/100,000 at >85 years; χ2 for trend = 44.4, p<0.001) and was higher among LTCF residents than among community residents (67.7/100,000 vs. 21.4/100,000; p<0.001). The overall case-fatality rate was 6.8 (8.7% LTCF vs. 6.4% community). Case-fatality rates increased as age increased (6.0% at 65–74 years, 6.8% at 75–84 years, and 8.2% at ≥85 years).
Comparison of LTCF residents and community-dwelling elderly persons with invasive GBS disease, Minnesota, 2003–2007*
The most common clinical presentation reported was bacteremia without focus (50.2%), followed by pneumonia (10.9%). LTCF residents (18.9%) were more likely than community residents (9.2%) to have pneumonia (p = 0.002) (). Blood (84.0%) was the most common site for isolation of GBS bacteria, followed by joint fluid (10.2%) and bone (3.3%). Other sites included peritoneal fluid (1.4%), pleural fluid (0.7%), and cerebrospinal fluid (0.4%).
Data on concurrent conditions were collected for 96 (75.6%) of 127 LTCF case-patients and 448 (75.2%) of 596 community case-patients. Of these, 176 (32.3%) had only 1 concurrent condition, 166 (30.5%) had 2 concurrent conditions, and 145 (26.6%) had >3 concurrent conditions. LTCF residents (94.8%) were more likely than community residents (88.4%) to have a documented concurrent condition (p = 0.06) (). Among case-patients with known concurrent condition status, 41% had diabetes mellitus and 30% had coronary artery disease; similar proportions were noted among LTCF and community case-patients. Congestive heart failure (26.0% vs. 15.0%, p = 0.009), stroke (13.5% vs. 5.1%, p = 0.003), and chronic obstructive pulmonary disease (15.6% vs. 8.3%, p = 0.026) were more common among LTCF residents. Cancer was more common among community residents (28.1% vs. 13.5%, p = 0.003) (). Cellulitis as a manifestation of invasive GBS disease was more likely in residents with diabetes than in those without diabetes (24.4% vs. 16.3%, p = 0.019).
GBS serotypes were obtained for 654 (90.5%) of 723 case-patients. Five serotypes, Ia (21.1%), Ib (11.0%) II (11.8%), III (11.3%), and V (35.0%), accounted for 94.6% of LTCF case-patients and 89.7% of community case-patients. Antimicrobial drug susceptibility data were obtained for 655 (90.6%) of 723 case-patients. All isolates were susceptible to penicillin. Susceptibility to erythromycin and clindamycin decreased during 2003–2007 (). Sixty percent of erythromycin-resistant, clindamycin-susceptible isolates had inducible clindamycin resistance as evidenced by a positive D test. During 2004–2005, 78% of erythromycin-resistant, clindamycin-susceptible isolates had inducible clindamycin resistance compared with 46% from 2006–2007 (p = 0.003). Serotype V was associated with higher rates of resistance than other serotypes to both erythromycin (46.7% vs. 27.9%, p<0.001) and clindamycin (28.4% vs. 12.9%, p<0.001). Serotype V was also associated with higher rates of inducible clindamycin resistance (88.6% vs. 43.4%, p<0.001).
Susceptibility of invasive GBS disease to erythromycin and clindamycin, Minnesota, 2003–2007*