| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Interdepartmental Division of Critical Care Medicine (Drs. Mehta, Grossman, and Stewart, and Mr. Hallett and Mr. Bowman) and Department of Microbiology (Drs. McGeer and Low), Mount Sinai Hospital and University of Toronto, Toronto, ON, Canada.
Correspondence to: Sangeeta Mehta, MD, Mount Sinai Hospital, 600 University Ave, Suite 18-216, Toronto, ON, Canada M5G 1X5; e-mail: geeta.mehta{at}utoronto.ca
Abstract
Study objectives: To describe the clinical features and outcome of patients with invasive group A streptococcal (GAS) infections admitted to the ICU.
Design: Prospective, population-based surveillance for invasive GAS infections was conducted in Ontario from January 1992 until June 2002. All 62 patients meeting clinical and/or histopathologic criteria for invasive GAS who were admitted to the ICUs of four university-affiliated hospitals in Toronto, Canada were included. Demographic and clinical information were obtained retrospectively by chart review. ICU morbidity data included the occurrence of organ dysfunction (renal, hepatic, coagulation, ARDS), treatment, and interventions such as hemodialysis and mechanical ventilation.
Measurements and results: ARDS developed in 34%, renal dysfunction developed in 55%, hepatic dysfunction developed in 64%, and coagulopathy developed in 69% of patients. A total of 56% of patients were treated with IV polyspecific IgG (IVIG), 81% were intubated and placed on mechanical ventilation, and 21% required renal replacement therapy. The median durations of ICU and hospital stay were 5.3 days and 15.0 days, respectively. The overall mortality was 40%. Mortality correlated directly with acute physiology and chronic health evaluation II score and the number of dysfunctional organs. Survivors were younger, had lower severity of illness scores, fewer dysfunctional organs, and were less likely to have shock or to receive treatment with vasopressors, mechanical ventilation, or pulmonary artery catheters. There was no association between the use of IVIG, surgical intervention, or clindamycin, and survival. Variables independently associated with mortality on multivariable analysis were the presence of coagulopathy (p = 0.0005) and liver dysfunction (p = 0.0123).
Conclusions: Patients with invasive GAS infection admitted to the ICU have a high mortality rate. In this group of patients, coagulopathy and liver failure were independently associated with mortality. We did not observe any association between the use of IVIG, surgical intervention, or clindamycin, and survival.
Key Words: ICU • ICU outcomes • group A streptococcus • necrotizing fasciitis • streptococcal toxic shock syndrome
Group A streptococcus (GAS) is a common pathogen of the throat and skin. It causes infections of varying severity, ranging from uncomplicated superficial to severe invasive. The past decade has witnessed a worldwide increase in severe invasive GAS infections, including streptococcal toxic shock syndrome (STSS) and necrotizing fasciitis (NF).1 These manifestations are rapidly progressive infections that are associated with high mortality rates despite prompt antimicrobial therapy.1 The mortality of STSS ranged from 33 to 81% in six population-based studies.234567
Although many population-based studies have described the morbidity and mortality of GAS, none have focused specifically on the outcome of patients requiring admission to the ICU. The purpose of this study was to describe the clinical course of patients with invasive GAS infections admitted to the ICU, and to identify risk factors for mortality in this population.
Materials and Methods
Patients
Active, population-based surveillance of invasive GAS infections in Ontario, Canada, has been conducted by our group since January 1992.3 All microbiology laboratories serving Ontario hospitals have a mandate to notify the study office whenever GAS is isolated from specimens from normally sterile sites. Annual audits are performed in all laboratories to ensure reporting accuracy. With this information, we conducted a chart review of all cases of invasive GAS admitted to the ICU of Mount Sinai Hospital, University Health Network, Sunnybrook and Women’s College Health Sciences Center, and St. Michael’s Hospital, all University of Toronto teaching hospitals. The study was approved by the Institutional Review Board of each hospital.
Clinical and demographic information was obtained from each patient’s chart, including age, gender, site of GAS infection, clinical manifestations, treatment, organ dysfunction, and outcome. Invasive GAS disease was defined as evidence of clinical infection in association with the isolation of GAS from a normally sterile body site or tissue. Soft-tissue infection was defined as infection associated with inflammation of skin or soft tissue but that excluded cervical adenitis, pharyngeal infections, and cutaneous inflammation overlying septic arthritis. Cases were classified as NF if histopathology demonstrated both necrosis of superficial fascia and polymorphonuclear infiltrate and edema of the reticular dermis, subcutaneous fat, and superficial fascia. In the absence of examined specimens, the diagnosis required the presence of gross fascial edema and necrosis detected at surgery, or frank cutaneous necrosis on physical examination if surgery was not performed. STSS was defined according to the criteria established by the Working Group on Severe Streptococcal Infections8—hypotension in combination with at least two of the following: acute renal failure, coagulation or liver abnormalities, rash, or NF.
Organ failure was defined as follows, using previously published definitions for organ failure in the setting of GAS infection.910 Acute renal failure was defined as a creatinine level > 176 µmol/L, or more than twice baseline for patients with chronic renal failure. Coagulopathy was defined as a platelet count < 100,000 x 109/L or evidence of disseminated intravascular coagulation (elevated partial thromboplastin time and positive d-dimer findings). Liver involvement was determined to be present if serum aspartate aminotransferase or serum alanine aminotransferase levels were more than twice the upper limit of normal, or twice baseline for patients with chronically elevated enzyme levels. ARDS was considered present if the patient had bilateral infiltrates on the chest radiograph, a ratio of PaO2 to the fraction of inspired oxygen < 200 mm Hg, and no clinical evidence of left ventricular failure. Cases were considered to be nosocomial if the disease was not present or incubating at the time of admission to the hospital. Clinical isolates were confirmed as GAS by use of standard methods. The APACHE (acute physiology and chronic health evaluation) II11 score was calculated for the first 24 h of admission, and the multiple organ dysfunction score (MODS)12 was calculated daily for the first 7 days.
To examine change in mortality and use of IVIG, clindamycin, and surgical debridement over time, the study period was divided into two 5-year blocks (January 1992 to July 1997, and August 1997 to June 2002). These two periods are of approximately equal duration, with equal numbers of patients in each period (n = 30 and n = 32, respectively).
Statistical Analysis
Data are presented as mean ± SD. Nonnormally distributed continuous variables are presented as median with interquartile range (IQR). Univariable associations between survival and demographic variables, organ dysfunction, site of infection, ICU interventions, and other important factors were examined by using univariable logistic regression for continuous variables, and Fisher Exact Test for categorical variables. Multivariable logistic regression was also performed. Among the variables that were significantly associated with death in the univariable analyses, we investigated the possibilities of multicollinearity among those independent variables. Since we did not have a large sample size, we only considered variables significantly associated with death, and those excluded from the multicollinearity analysis, to be included into the multivariable logistic regression model building process. A three-step procedure was used.13 First, stepwise regression was performed, with all variables entered and remaining in the model to build the stepwise sequence. Second, Akaike information criterion/Schwarz information criterion optimal and suboptimal models for the stepwise sequence were found. Third, best subset selection was applied to the sample sizes that correspond to Akaike information criterion/Schwarz information criterion-optimal and nearly optimal models. The corresponding p values, odds ratio estimates, and their 95% confidence limits from likelihood methods were reported. Statistical analysis software (SAS version 8.02 for Windows; SAS Institute; Cary, NC) was used to analyze the data; p < 0.05 was considered statistically significant.
Results
Between January 1992 and June 2002, 62 patients with invasive GAS disease were admitted to the ICU in one of the four hospitals. Baseline demographic data are presented in Table 1 . More than 60% (n = 40) of patients initially presented with skin or soft-tissue infections, and 27 of these met the criteria for NF. Pneumonia was the second-most-common site of infection at presentation. Sixty-eight percent of patients had blood culture results positive for GAS, four of whom had no obvious source of infection. At presentation, approximately 50% of patients had a history of chronic disease, including coronary artery disease, congestive heart failure, and pulmonary, renal, or liver disease.
|
|
The overall mortality rate was 40% (n = 25). The median time to death from ICU admission was 6.1 days. Mortality correlated directly with APACHE II scores (Fig 1 ; p < 0.0001) and with the number of organ failures (Fig 2 ; p < 0.0001). Patients with four organ failures had a mortality of 83%. Mortality in patients with and without confirmed GAS bacteremia was 46% and 39%, respectively (p = not significant [NS]). Fifty-five percent of patients (n = 34) met the criteria for STSS. Mortality rates in patients with and without STSS were 68% and 8%, respectively (p < 0.0001).
|
|
|
|
In this retrospective case series, we found, as expected, that ICU patients with invasive GAS infections have significant morbidity and mortality. However, we were not able to show that IVIG has a beneficial effect in the treatment of these patients, in that mortality was similar in patients who did or did not receive IVIG.
The two most severe manifestations of invasive infections caused by GAS are NF and STSS.14 GAS express several virulence factors, both cell bound and secreted, and a pivotal role for superantigens in mediating the systemic effects of STSS has been shown.1516 Protective humoral immunity to GAS virulence factors is important in preventing invasive disease.1718192021 Patients with invasive GAS disease show significantly lower serum levels of protective antibodies against M-protein and superantigens than sera from noninvasive cases.17181920 These findings provide evidence that lack of protective humoral immunity against GAS virulence factors contributes to susceptibility to invasive infection.
IVIG is suggested as a potential adjunctive therapy for invasive GAS diseases due to its ability to neutralize a wide variety of superantigens and to facilitate opsonization of streptococci.22 However, support for clinical efficacy of IVIG in severe GAS infections is weak, consisting of a case series,23 a case-control study,24 and one small randomized trial.25 In the case-control study, survival was higher in 21 patients with STSS treated with IVIG than in 32 patients who did not receive IVIG (67% vs 34%, p = 0.02).24 The efficacy of IVIG in STSS was evaluated in a multicenter, randomized, placebo-controlled trial25 that was terminated after the enrollment of 21 patients because of slow recruitment. Although the trial was underpowered, mortality was 3.6 times higher in the placebo group compared with the IVIG group (p = NS), and there was a significant decrease in sepsis-related organ failure scores at days 2 and 3. In these trials, IVIG doses of up to 2 g/d for 1 to 5 days have been shown to increase superantigen neutralizing activity.
Our patients had significant morbidity. Reported rates of organ dysfunction in non-ICU case series of invasive GAS vary widely but are understandably lower than those reported in the current study. Reported ARDS frequencies range from 14%9 to 40%.26 Acute renal failure has been reported in approximately 40%,926 with the need for renal replacement in 5%.26 Kaul et al9 observed that coagulopathy developed in 29% of patients with STSS, and 28% had abnormalities in liver function test results.
Mortality in the current study, because we included only patients admitted to the ICU, differs from previous studies that have reported mortality in all cases of invasive GAS. Population-based series from Montreal,6 Ontario,3 Israel,27 and the Netherlands7 have observed mortality rates between 14% and 18% for invasive GAS. However, two studies have reported mortality rates similar to ours. Kaul et al24 reported 34% mortality in patients with GAS NF, and Mulla28 reported a mortality rate of 38.7% in a subset of 75 patients with invasive GAS admitted to Florida ICUs. The latter was a brief report that did not describe patient characteristics; thus, it is difficult to compare their patient population with ours.
Several investigators have questioned whether patients who are bacteremic with GAS have a worse outcome than those without bacteremia. Reported rates of bacteremia range from 46 to 85%.3910262729 In our study, 68% of the patients had GAS bacteremia, and nonsurvivors were more likely to be bacteremic; however, the difference was NS. Similarly, other studies2627 have observed higher mortality in bacteremic vs nonbacteremic patients. The consistently higher case fatality rate in bacteremic patients in several series suggests that this difference is real; however, it is too small to be clinically useful.
On multivariable analysis, we identified the occurrence of liver dysfunction and coagulopathy as factors significantly associated with mortality. In contrast, Kaul et al24 identified lower APACHE II score and IVIG therapy as the only two variables significantly associated with survival. In a prospective, population-based survey in Ontario, Canada, the same group9 identified older age, the presence of hypotension, and bacteremia as risk factors for mortality. In the Israeli series,27 nonsurvivors were older and were more likely to have bacteremia, STSS, and one or more comorbidities.
GAS remains universally susceptible to ß-lactams; thus, penicillin is the first choice for treatment. Clindamycin has been proposed as a useful adjunct for the treatment of severe GAS infections because it interferes with the synthesis of toxins/superantigens, and may reduce toxin production before ß-lactams kill bacterial cells. Indeed, Mulla et al29 observed that treatment with clindamycin reduced mortality in patients with NF. However, in the current study, we did not observe any association between the use of clindamycin and outcome.
This study has a number of limitations. It is a retrospective study, with no predetermined criteria for ICU admission. In addition, since we are a referral center for GAS infections, our cases may not be representative of population-based disease. The relatively small sample size of the study also means that the power to detect and label as statistically significant risk factors for mortality is limited, and the absence of association in our study does not mean that risk factors or aspects of therapy are not important. Finally, given the observational study design, our study neither supports nor refutes the value of IVIG, clindamycin, or surgical intervention in the treatment of invasive GAS infections.
Conclusions
Invasive GAS in adults is a disease that may have a fulminant course. Management of severe GAS infection remains supportive, and the mortality associated with invasive GAS has remained unchanged over the last 10 years. IVIG is currently recommended for patients with STSS; however, this recommendation is based on a single case-control study.24 In this retrospective case series, we were not able to show that IVIG has a beneficial effect. In view of the significant morbidity and mortality of invasive GAS infections, further studies are needed to define the role of IVIG in its management.
Footnotes
Abbreviations: APACHE = acute physiology and chronic health evaluation; GAS = group A streptococcus; IVIG = IV polyspecific IgG; IQR = interquartile range; MODS = multiple organ dysfunction score; NF = necrotizing fasciitis; NS = not significant; STSS = streptococcal toxic shock syndrome
Dr. Mehta has received a peer-reviewed grant from Bayer Healthcare- Biological Products Division to study IVIG utilization in Toronto hospitals. Dr Stewart has received honoraria from Bayer Healthcare- Biological Products Division for his role as an advisor.
Received for publication November 1, 2004. Accepted for publication November 18, 2005.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
