During 2008–2019, a total of 483 cases of iGAS disease were reported among Idaho residents. Annual disease incidence per 100,000 persons increased from 1.04 to 4.76 during 2008–2019 (Figure 1). Case numbers were highest in January and lowest in August with similar patterns across years (Figure 2; Appendix Figure 1). We identified no changes in the surveillance system that might have led to increased case reporting. Overall, 52.8% of cases occurred among men; median age of case-patients was 62 years (Table 1). Data concerning race and ethnicity were available for 444 (91.9%) patients, most of whom were white, non-Hispanic persons (87.8%), followed by Hispanic (5.0%) and American Indian or Alaska Native persons (4.1%). For comparison, the 2018 population of Idaho was estimated to be 83.1% white non-Hispanic, 12.7% Hispanic, and 2.0% American Indian or Alaska Native (22). The average annual incidence per 100,000 persons was 2.39 during 2008–2019; incidence was 2.55 among white non-Hispanic, 1.03 among Hispanic, and 5.19 among American Indian and Alaska Native persons. Data concerning residence type were available for 428 (88.6%) patients; most (349, 81.5%) resided in a private residence, followed by 63 (14.7%) in a LTCF or nursing home.

Mean annual incidence per 100,000 persons was 0.94 during the 2008–2013 baseline period (n = 89 cases) and 3.84 during 2014–2019 (n = 394 cases). Patient demographics were similar between the periods (Table 1); we included odds ratios in risk factor analysis (Table 2). Disease incidence increased between periods among all age groups (Figure 3). The mean annual age-standardized incidence per 100,000 persons increased from 1.0 (95% CI 0.5–1.5) during 2008–2013 to 3.7 (95% CI 2.8–4.6) for 2014–2019. iGAS incidence increased in all 7 of Idaho’s public health districts (Appendix Figure 2).

Medical records were available for 383/483 (79.3%) cases, 62/89 (70%) during the baseline 2008–2013 period and 321/394 (81.5%) during 2014–2019. Information concerning deaths was available for 464 (96.0%) patients, 55 (11.9%) of whom died. The case-fatality rate was 14/84 (16.7%) for the baseline period and 41/380 (10.4%) during 2014–2019 (p = 0.132). Of 471 patients with data on hospitalization status, 441 (93.6%) were hospitalized. The proportion of patients hospitalized was slightly lower during 2008–2013 (88.4%, 76/86) compared with 94.8% (365/385) during 2014–2019 (p = 0.027). Data on length of hospital stay were available for 385 hospitalized patients. Median stay was 5 days (IQR 4–9) for all hospitalized patients: 5 days (IQR 3–8) for 58 patients hospitalized during 2008–2013 compared with 6 days (IQR 4–9; p = 0.118) for 327 patients hospitalized during 2014–2019.

Cellulitis was the most common clinical syndrome, reported in 41.4% of cases, followed by bacteremia without focus (34.2%) and pneumonia (16.8%) (Table 3). STSS, a rare but severe syndrome caused by GAS infection, was identified in 25 cases, all during 2014–2019 (p = 0.02). Toxic shock syndrome is a reportable disease in Idaho; however, 11 (44%) STSS cases were identified only retrospectively through medical record review. Ages of STSS patients ranged from 10 months to 82 years; 6/22 (27%) died (data missing for 3 patients). We observed no other differences in clinical syndromes between periods. GAS was cultured from blood in 92.9% (442/476) of cases, with no differences over time: 74/82 (90.2%) for the baseline period compared with 368/394 (93.4%; p = 0.31) during 2014–2019. Data on postpartum status were available for 47/50 women 15–44 years of age, 10/47 (21%) of whom were postpartum, with no difference between periods: 2/9 (22%) during the baseline period compared with 8/38 (21%; p = 0.94) during 2014–2019.

emm typing was conducted on bacterial isolates from 194 (40.2%) iGAS cases, 2/89 (2.3%) during 2008–2013 and 192/394 (48.7%) during 2014–2019. In total, we identified 38 different emm types; the most common were types 1 (n = 26, 13%), 12 (n = 25, 13%), 28 (n = 23, 12%), 11 (n = 15, 8%), and 4 (n = 15, 8%) (Figure 4; Appendix Table 1). emm typing results were available for 14/25 (56%) STSS cases, from which 10 emm types were observed; types 1 (n = 3), 12 (n = 2), and 1.25 (n = 2) were identified in >1 patient.

Two outbreaks were previously identified on the basis of epidemiologic information and emm types. During July–September 2016, an outbreak of 5 cases of iGAS caused by emm59 occurred among residents of a single county. During 2014–2016, an outbreak of iGAS occurred among residents of a LTCF; 9 cases were emm11, and 4 cases did not have emm typing conducted. In addition, a household cluster of 2 cases, both associated with injection drug use, occurred in May 2015, but emm typing was not conducted. In total, 16/394 (4.1%) of cases during 2014–2019 were associated with a cluster or outbreak.

Information concerning underlying medical conditions was available for 432 (90.2%) patients, 69/89 (78%) during 2008–2013 and 363/394 (92.1%) during 2014–2019. Of these patients, 73.8% had ≥1 underlying condition; diabetes (41.2%), heart disease (26.9%), and obesity (22.0%) were the most common (Table 4). No patients had HIV infection. Overall, 201/412 (47.8%) patients had skin injuries, and nonsurgical wounds were most frequently reported. Injection drug use was reported for 8/386 (2.1%) and methamphetamine use for 6/386 (1.6%) patients, all during the 2014–2019 period. Data for other risk factors (GAS pharyngitis, household member with GAS infection, influenza infection) were available for 389 patients, and GAS pharyngitis was identified in 8.0% of patients. In regression analyses, we observed no associations between demographic or risk factors and the higher-incidence 2014–2019 period (Table 2). Injection drug use had an adjusted odds ratio of 3.2; however, the limited number of observations yielded wide 95% CIs.

Using statewide reportable disease surveillance data supplemented by information from medical record review, we investigated the epidemiology of iGAS in Idaho over a 12-year period, during which incidence increased ≈>4-fold. The ABCs program, which does not include data from Idaho, identified a similar increase in nationwide iGAS incidence per 100,000 persons, from 3.69 in 2008 to 7.63 in 2019 (1,12,23). In Canada, incidence per 100,000 persons rose from 4.42 in 2008 to 8.12 in 2019 (https://diseases.canada.ca/notifiable/charts).

In Idaho, average annual iGAS incidence during our study was 2-fold as high among American Indian or Alaska Native persons compared with white non-Hispanic persons. Higher iGAS incidence among indigenous compared with nonindigenous populations has been reported in multiple settings, including the United States, Australia, and New Zealand (24). In Alberta, Canada, iGAS incidence was 6-fold as high and increasing more rapidly among First Nation compared with non-First Nation populations, rising to 52.2 cases per 100,000 persons in 2017 (25). Neither the proportion of cases occurring among American Indian and Alaska Native persons nor the racial and ethnic distribution of Idaho’s population changed substantially during our study period (Appendix Table 2).

Approximately three quarters of iGAS patients had >1 underlying medical condition; diabetes, obesity, and skin injuries were common, emphasizing the need for diabetes management and wound care to reduce the risk for iGAS associated with these conditions (4,5). The high proportion of cases associated with skin injury or infection is consistent with national data; in 2019, a total of 44.7% of iGAS case-patients had cellulitis (1). Two outbreaks and a household cluster were identified through routine surveillance, all during 2014–2019; emm typing made outbreak detection easier. Because the state public health laboratory only began emm typing in 2014, earlier outbreaks might not have been detected. We recommend that states with available resources conduct emm typing to enhance surveillance and outbreak detection.

The incidence of iGAS in Idaho, 4.76 cases per 100,000 persons in 2019, is lower than the national incidence rate, 7.63 per 100,000 persons in 2019, possibly because Idaho iGAS estimates are based on passive data reporting, whereas national estimates are based on data from the ABCs active surveillance system. Other factors related to demographics or prevalence of risk factors might also contribute to lower Idaho compared with national iGAS incidence estimates. Increased iGAS rates across all age groups and in all geographic regions of Idaho during 2008–2019 present cause for concern. We identified no changes in surveillance procedures that might have led to increased case reporting. However, we did not assess potential changes in clinical practice (e.g., the number of blood cultures ordered by emergency departments), which might have affected case detection.

We identified a rise in STSS incidence: 25 cases occurred during 2014–2019, compared with none during 2008–2013, although these numbers do not account for the increase in overall iGAS cases. We found evidence of STSS underreporting, as has been reported in other studies (26). Cases of STSS from the baseline period might have been missed because medical records were only available for 70% of iGAS patients. We observed 10 different emm types among STSS cases that had an isolate submitted for typing, indicating a lack of clonality. Of note, a study in the Netherlands identified a temporal association between STSS and influenza A virus (27). However, STSS cases in Idaho occurred throughout the year, without seasonal patterns that might have been associated with influenza.

emm1 was the most common emm type identified in Idaho and also the most common cause of iGAS nationally during 2008–2019 (1,12,23). The diversity of emm types observed in Idaho was consistent with reports from high-income countries in North America and Europe (28–30). Although no vaccines targeting GAS are available, the 30-valent M protein–based strep A vaccine, one of the few vaccines under development in phase 1 or 2 clinical trials, would potentially cover 167/194 (86.1%) emm types from Idaho iGAS cases, and possibly more through cross-protection within emm clusters based on M protein structure (31–33).

We identified no associations between risk factors and increased incidence during 2014–2019. In New Mexico, increases in iGAS incidence were linked to rises in cases among persons experiencing homelessness and persons injecting drugs (17). Similarly, an analysis of ABCs national surveillance data concluded that injection drug use and homelessness likely contribute to increasing iGAS incidence in the United States, noting that particular emm types are increasing among this patient population (7). Factors responsible for increased incidence of iGAS during 2003–2017 in Alberta, Canada, were not clear, and multiple risk factors likely contributed, particularly alcohol abuse and drug use (13). In Idaho, only 8 patients reported injection drug use and 9 experiencing homelessness. Although these conditions might be underreported, homelessness and injection drug use do not appear to be driving the increase in iGAS in Idaho, although they might be contributing factors. We found no evidence that alcohol abuse was contributing to the increase in iGAS in Idaho.

Our results suggest that iGAS is increasing among the general population of Idaho and not limited to a particular age group or associated with an individual risk factor. The reasons for increases in reported incidence in Idaho might be multifactorial and involve factors not assessed in this study, such as expansion of >1 emm types and improvements in collection of diagnostic specimens by clinicians and passive reporting by laboratories. One limitation of our study is that we did not directly assess whether changes in the distribution of risk factors in the underlying population, such as the proportion of adults with skin injuries, could have led to increases in iGAS. For certain characteristics and risk factors, Idaho population data on estimated prevalence or proportion were unavailable (Appendix Table 2). For most factors, only minor changes over time were observed, and we did not identify any changes considered potentially responsible for the increase in iGAS incidence. Another limitation to the risk factor analysis was the small sample size during the baseline period—only 89 cases, some missing data. A third key limitation to our study was the lack of emm typing data from the 2008–2013 baseline period, because only 2% of cases had emm typing conducted, compared with 49% during 2014–2019. Therefore, we could not assess potential shifts in emm types, which might have contributed to increasing rates of iGAS. In Ireland, an increase in iGAS during 2012–2015 coincided with an upsurge in emm3 (34). In the United States, expansion over time of select emm types, including 11, 77, and 92, was linked to increasing prevalence of antimicrobial drug–resistant iGAS infections (35). Factors that contribute to the emergence of emm types are not well understood but might include genetic changes in bacteria and acquisition of new virulence factors, as well as variations in population immunity (30). emm59, which emerged initially in Canada and more recently has been found in the southwestern United States, was the most common emm type in Idaho in 2014; all 5 cases were associated with outbreaks, but emm59 declined during subsequent years (19,36).

In conclusion, in Idaho, iGAS is an urgent and consequential clinical and public health concern because of its severity, high case-fatality rate, and statewide increases. The absence of an identified risk factor contributing to increasing iGAS incidence in the general population suggests a potential role for vaccination as a preventive measure. Ongoing surveillance of state-level iGAS cases would help identify outbreaks, track regional trends in incidence, and monitor circulating emm types.