We enrolled patients during January 1, 2005–December 31, 2006. Patients were included if they resided in metropolitan Auckland and had a GAS isolate cultured from a previously sterile body cavity. Patients with STSS were included in accordance with the consensus definition (19); STSS also was the diagnosis for patients who were dead on arrival or who died within 48 hours after illness onset and for whom laboratory data were insufficient in accordance with the methodology of Davies et al. (1). NF was defined as tissue necrosis diagnosed by histopathologic examination or by the treating surgeon during surgical debridement. Patients could have had >1 diagnosis, with the exception of bacteremia without a source. Clinical syndromes, such as skin or soft tissue infection, had to be accompanied by recovery of an isolate from a normally sterile site or specimen, such as blood, to meet the case definition. Nosocomial infection was defined as GAS infection in patients who had been hospitalized for >72 hours. Invasive GAS infection was defined as postpartum if it occurred in a woman who was pregnant or <30 days after delivery or who had clinician-defined puerperal fever, chorioamnionitis, or a septic abortion. Women from whom GAS was isolated from amniotic fluid or placenta alone were excluded (20). Our study was approved by the regional ethics committee and each hospital’s research committee and Maori research committee.

Data were collected from the microbiology laboratories serving all 3 Auckland regional District Health Board (DHB) hospitals, i.e., Auckland City Hospital and Starship Children’s Hospital (Auckland DHB); Middlemore Hospital, which includes Kidz First Children’s Hospital (Counties Manukau DHB); and North Shore Hospital and Waitakere Hospital (Waitemata DHB). All Auckland residents with serious medical illness would attend 1 of these hospitals.

Auckland (2006 population: 1,387,780), New Zealand’s largest city, comprises one third of the country’s population and is the country’s most ethnically diverse city. In 2006, 19.0% of residents self-identified as Asian, 14.4% as Pacific Islander, 11.1% as indigenous Maori, and 56.5% as European. The climate is temperate, with summer occurring during December through March. We used New Zealand birth data for infants <1 year of age and customized New Zealand census charts for DHBs as denominators.

We obtained demographic and clinical features by reviewing medical charts and electronic documents. To ensure complete surveillance, we requested International Classification of Diseases, 10th Revision, diagnoses from DHB data managers. We contacted the regional coroner and forensic pathologist to seek out records of deaths (including deaths in the community) caused by GAS infection and scrutinized intensive care unit (ICU) data for diagnoses of shock from GAS, STSS, or NF. Disease severity was determined by length of stay, ICU admission, and use of surgical and medical procedures.

We assigned each invasive GAS infection in the Auckland region a deprivation score by using the New Zealand Deprivation Index 2006 (www.moh.govt.nz). This index measures socioeconomic status (SES) in small areas according to 9 variables (income, income assistance, education, access to a car and phone, household crowding, employment, single-parent family, housing rented or owned).

β-hemolytic colonies on blood agar were typed as Lancefield group A by using commercially available latex agglutination kits (Pro-Lab Diagnostics, Austin, TX, USA). Group A isolates were sent to Environmental Science and Research Laboratory (Wellington, New Zealand) for emm typing by using established procedures (21). Concordance between emm types and M serotypes has been established (21). Antimicrobial drug sensitivities were determined by routine methods (22).

We used emm typing to estimate the proportion of cases and deaths caused by emm types in the proposed 26-valent vaccine. These emm GAS types are 1, 2, 3, 5, 6, 11, 12, 13, 14, 18, 19, 22, 24, 28, 29, 33, 43, 59, 75, 76, 77, 89, 92, 94, 101, and 114 (14). We then calculated potential vaccine efficacy in the most at-risk Auckland populations: persons <5 years of age and >65 years of age.

During the 24-month study period, we identified 333 patients who potentially had invasive GAS infections. Of these, we excluded 118 who did not fulfill the inclusion criteria. The most common reasons for exclusion were isolation from a nonsterile site or residence outside metropolitan Auckland at the time of diagnosis. Using the electronic discharge summaries based on International Classification of Diseases, 10th Revision, coding, we identified and included 10 (4%) additional cases that fit the case definition.

The 225 patients were from all ethnic groups: European (77 [34%] patients), Maori (69 [31%]), Pacific Islanders (70 [31%]), and other ethnicities (7 [3%]). For 2 patients, no information was available about ethnicity. For the 225 patients, median age was 53 years (range 0–97 years), and 119 (53%) patients were male. Ethnic disparities, although notable in the extremes of life, did not differ significantly by age (Figure 1; Table 1). The 198 patients with invasive GAS infection for whom SES information was available were more likely to originate in areas designated by the New Zealand Deprivation Index 2006 as lower SES areas than in higher SES areas (Table 2). Forty-nine percent of case-patients were from the lowest SES quintile.

Twenty-two patients died, for an overall case-fatality rate (CFR) of 10% (Figure 1, Table 3). Fourteen of these patients died within 72 hours after hospital admission. Three infants (one 2 months of age and two 4 months of age) who died in the community had STSS. One death previously had been attributed to sudden infant death syndrome.

The median age of patients who died was 62 years (range 2 months–86 years). Eighteen adults who died had multiple concurrent illnesses. The highest CFR (31%) was for infants (a total of 4 deaths in three 4-month-old infants and one 2-month-old infant); these were the only deaths among children <15 years of age.

All infants who died had GAS-positive blood cultures. One who died in the community also had GAS-positive cerebrospinal fluid. Three of the 4 deaths occurred in the community and are attributed to STSS. The illness of the fourth (hospitalized) infant also met the criteria for STSS. Bronchopneumonia was found at post-mortem examination in 2 infants (1 hospitalized, 1 in the community). Two of the infants who died in the community had additional pathogens isolated from postmortem blood cultures (Staphylococcus aureus in both cases and Streptococcus pneumoniae and viridans streptococci in 1 each) but no gram-negative organisms.

The most common clinical feature was skin and soft tissue infection (97/225; 43%) (Table 3). Of the 30 patients with STSS, 26 (87%) had an underlying condition before the onset of acute GAS disease. Median age at STSS occurrence was 57 years (range 2 months–86 years). Six cases occurred in children <5 years of age. Empyema (4 cases; p<0.0001) and brain abscess (2 cases; p = 0.0011) occurred more frequently in children <14 years of age than in adults. The incidence of bacteremia with no focus of infection was 1.4 cases per 100,000 persons per year overall, but 3.7 per 100,000 for children <5 years of age (n = 7).

Seven cases of GAS postpartum infection were recorded for women 15–49 years of age, for a rate of 0.16 cases per 1,000 live-born infants (Maori and Pacific Islander, 0.21 cases/1,000 live-born infants). No deaths occurred in this group. We also identified 3 premature neonates with invasive GAS disease unrelated to cases in adults; 1 infection was nosocomially acquired.

Of the 223 patients for whom data were available, 58 (26%) had no underlying condition or other risk factor, 114 (51%) had 1 or 2 risk factors, and 64 (28%) had >3 risk factors. In the >15-years age group, 67 (35%) had heart disease, 60 (32% [23 Maori, 25 Pacific Islanders]) had diabetes, and 21% had either renal disease (39 persons) or lung disease (40 persons). Cigarette smoking was the most common nondisease-related risk factor (56 [30%] of 189 persons >15 years of age).

GAS was most frequently isolated from peripheral blood cultures (184 [82%]). Other sources were surgical specimen (37 isolates), tissue specimen (18), joint aspirate (16), pus aspirate (12), catheter blood culture (6), peritoneal aspirate (2), cerebrospinal fluid (2), and postmortem blood (3).

Of the 225 cases, 205 (91%) GAS isolates were available for emm typing (Figure 2). Seventy (34%) of 205 cases had an emm type that was contained in the 26-valent vaccine. The proposed 26-valent vaccine could prevent 30% of GAS invasive cases in children <5 years of age and 15% of cases in persons >65 years of age (Table 4).

Of the 225 isolates, 1 (0.4%) was resistant to erythromycin and 1 (0.4%) had intermediate sensitivity to erythromycin. Three (1.3%) were resistant to clindamycin.

Hospitalization was required for 222 patients (3 deaths occurred in the community). Length of stay was >10 days for 105 (47%); mean length of stay was 15.9 days (range 1–153 days). Thirty-eight (17%) required ICU admission (mean length of stay 4.5 days; range 1–9 days); maximum length of stay was 19 days. Nosocomial infection was responsible for 12 (5%) of the 225 cases. Seventy-five (33%) patients required at least 1 surgical procedure, predominantly drainage, débridement, or washouts. One patient (2 years of age) with STSS required intravenous immunoglobulin.