Background. Nosocomial transmission of group A streptococcus (GAS) has been well described. Instances resulting in fulminant disease among health care workers have not been described. Contact and droplet precautions have been advised to minimize the risk of nosocomial transmission. We aimed to determine whether a case of invasive GAS pneumonia and streptococcal toxic shock syndrome in a respiratory therapist was acquired as a result of caring for a patient with GAS necrotizing fasciitis. Contacts of these patients were screened to determine if they were the reservoir of the GAS. Genetic testing to confirm clonal transmission was conducted.
Methods. Contacts of the patients with GAS infection were screened using questionnaires and testing of pharyngeal specimens. Specimens from patients and carriers of GAS who were identified during screening were subjected to pulsed-field gel electrophoresis and emm gene typing.
Results. We identified 705 contacts of the 2 patients; all contacts had oropharyngeal specimens collected for culture. Only the index patient and the respiratory therapist yielded identical pulse-field gel electrophoresis GAS isolates, with clonality indicated by emm typing.
Conclusions. Nosocomial transmission of GAS from a patient to a health care worker who developed streptococcal toxic shock syndrome may have occurred after the index patient had received 48 h of antibiotic therapy and despite placement in contact isolation. Isolation guidelines for patients with severe GAS infection may need to be reviewed.
Streptococcus pyogenes (group A streptococcus [GAS]) is primarily a community-acquired pathogen causing 9000–11,000 cases of invasive disease in the United States each year [1]. The reemergence of severe toxic streptococcal disease has coincided with an epidemiological shift in GAS strains; an increased prevalence of certain M protein serotypes, such as M1 and M3; and an increase in the number of strains producing streptococcal pyrogenic exotoxin A and protease [2, 3]. Cases of nosocomial GAS transmission in the hospital setting have resulted in lethal patient outcomes. Guidelines for preventing nosocomial transmission of GAS were updated in 2007; these recommend that patients with invasive or severe streptococcal disease remain in droplet and contact precaution for 48 h after effective therapy has been initiated [4]. Although transmission of severe invasive GAS disease from an asymptomatic health care worker (HCW) to a patient has been described, the reverse scenario has not been reported. We describe the apparent nosocomial transmission of GAS from an infected patient to a respiratory therapist that resulted in fulminate invasive disease.
The index patient was a homeless 34-year-old Native American female with a history of alcoholism who presented to the emergency department on 9 February 2005 (at 8:43 pm) complaining of weakness, arthralgias, and vomiting, with severe left leg, hip, and back pain. The leg pain developed after a vaguely recalled fall during the previous week. A moderate cough with greenish sputum was reported without a sore throat for a few days before admission to the hospital. She was actively menstruating but did not use tampons. There was no significant medical history or allergies. The patient's blood pressure was 62/40 mm Hg, her pulse was 141 beats/min, her respiratory rate was 22 breaths/min, and her temperature was 38.1°C. She was acutely ill and in distress because of pain. The head, neck, cardiopulmary, and abdominal examination findings were unremarkable. She had mottled, cool extremities with an ecchymotic and tender left thigh without crepitus or bullae. Chest, pelvic, and left femur radiograph findings were normal, as were urine analysis findings. No sputum or throat swab samples were obtained for culture. The patient's WBC count was 9600 cells/mm3(with 25% neutrophils and 24% band forms), her creatine phosphokinase level was 4294 U/L, and her creatinine level was 2.4 mg/dL. The total bilirubin level was 0.8 mg/dL, and levels of hepatic transaminases were moderately elevated. Blood cultures were performed. She remained hypotensive despite fluid infusions with >5 L of saline and became less responsive after receiving morphine therapy. At 9:26 pm, the patient received 2 g of ceftriaxone intravenously (IV), underwent endotracheal intubation for mechanical ventilation, and received vasopressor agents. A surgical consultant performed a bedside thigh muscle biopsy at 11:30 pmand concluded that there was no sign of infection. The consultant sent a biopsy sample for microbiologic studies and recommended adding aztreonam (2 g IV every 8 h) and clindamycin (900 mg IV every 8 h) to the treatment regimen. The patient was transferred to an intensive care unit and was seen by an infectious diseases consultant the next day at 10:00 am. The infectious diseases consultant observed mottling of both legs, primarily proximally. The left thigh was indurated 50% larger than the right, with extension of hemorrhagic and violatious mottling in the proximal left thigh, inguinal, and mons area. A wound specimen for Gram stain and culture that was obtained from the biopsy lesion showed few neutrophils and gram-positive cocci. Surgical debridement was advised. Clindamycin and ceftriaxone therapy was continued. Blood and wound specimens were positive for GAS within 24 h after obtainment. Approximately 16 h after presentation, the patient had left hip disarticulation caused by massive thigh myonecrosis. Specimens obtained intraoperatively were still positive for GAS. Contact precautions (gloves and gowns when entering the patient's room) were implemented at admission to the intensive care unit and were maintained for the course of therapy. Postoperatively, the patient received IV immunoglobulin (50 g every 6 h for 4 doses) and required several more debridement procedures of the lower abdomen and pubic area. She improved and was extubated on hospital day 12. After skin graft and flap procedures, the patient was discharged to a rehabilitation hospital on day 52.
The second case occurred in a 53-year-old African American male respiratory therapist who presented to the same hospital emergency department on 15 February 2005 (at 8:55 pm) with nearly a week of cough; dyspnea, which acutely worsened on the day of presentation; pleuritic chest pain; and fever. He had a mild sore throat. He worked at the hospital where the index patient was admitted, including in the intensive care unit. He did not use alcohol, illicit drugs, or tobacco and did not have a relevant medical history. His wife and son had mild cold symptoms for several days before the HCW's illness. At presentation, his blood pressure was 115/70 mm Hg, his pulse was 115 beats/min, and his temperature was 35.3°C (orally). His respiratory rate was 22 breaths/min, with gasping, and the room air oxygen saturation was 85%. There was no pharyngitis or exudate. The chest had left rales, but the remainder of the examination findings were unremarkable. A portable chest radiograph revealed a left lower lobe infiltrate. The peripheral WBC count was 12,900 cells/mm3(with 57% neutrophils and 21% bands), the serum creatinine level was 3.1 mg/dL, and the total bilirubin level was 1.8 mg/dL; transaminase levels were normal, and the alkaline phosphatase level was 170 U/L. A sputum specimen showed neutrophils and gram-positive cocci in chains. Blood cultures were performed, and ceftriaxone (1 g IV) and azithromycin (500 mg IV) therapy was administered. The HCW's systolic blood pressure decreased to 95 mm Hg despite crystalloids. Respiratory failure ensued, requiring endotracheal intubation and mechanical ventilation at ∼3 h after presentation. He was admitted to the intensive care unit and received an additional 1 g of IV ceftriaxone and 1 g of vancomycin. The blood and sputum culture grew GAS isolates that were susceptible to macrolides and penicillin. Complications included ventilator-associated pneumonia and gastrointestinal bleeding. While still receiving mechanical ventilation, the HCW was transferred to a rehabilitation facility 50 days after his presentation. In 2007, he returned to work full time as a respiratory therapist.
Because these 2 cases of invasive GAS disease occurred so closely in time to one another (within a week), the isolates from each of the cases were sent to the Arizona Department of Health Services (ADHS; Phoenix) for pulsed-field gel electrophoresis (PFGE). The investigation began when infection control was notified by ADHS on 1 March 2005 that the isolates were identical by PFGE, which indicated that these isolates were genetically related. Subsequently, the isolates were sent to the Centers for Disease Control and Prevention (CDC; Atlanta, GA) for T typing and M protein gene (emm) typing.
Investigation was performed of the initial contact between the infected HCW and the index patient, which occurred ∼51 h after the admission of the index patient, on 12 February 2005 (at 3:00 am) for ∼30 min. The respiratory therapist also worked with the index patient for the entire shift on 13 February 2005, ∼96 h after admission of the index patient. On 14 February, the HCW arrived at work at 6:00 pmbut became increasingly ill and went home early (4 h after starting his shift).
Discussion was initiated with the CDC and ADHS to begin looking for asymptomatic carriers among the HCWs at the hospital to try to link these 2 cases. A contact was identified as an asymptomatic or symptomatic family member or HCW who had had direct contact (defined as touching the patient or the bed of the patient or care provided within 90 cm of the index patient or the respiratory therapist). A questionnaire was given to all contacts to illicit symptoms of sore throat, fever, open draining sores, swollen tonsils with exudates, rash (for >1 day), impetigo, swollen lymph nodes, or arthralgias within the previous 7 days. Throat cultures and rapid streptococcus screens were performed for anyone with the aforementioned symptoms.
Throat swab samples were collected in a standard culture tube and were sent to the ADHS for culture. GAS was identified by means of standard methods [5]. The staff at the ADHS performed all PFGE analyses on positive throat culture and blood culture samples. PFGE was performed using the Bio-Rad Chef Mapper system, and the preset Staphylococcus setting was used with the buffer temperature at 14°C for 18.5 h. The gel was stained with ethidium bromide and was photographed. Isolates with identical DNA fingerprints by PFGE (i.e., those that did not differ by >1 band) were sent to the Streptococcus Laboratory at the CDC Respiratory Diseases Branch for sequencing of the M protein gene (emm bv typing). M typing and T agglutination pattern identification were performed using standard methods [6]. No additional testing for pyrogenic exotoxins or other virulence factors was performed.
Seven hundred five HCWs were screened by throat culture. Fourteen HCWs were positive for GAS. Ten were asymptomatic carriers. Four were symptomatic with fever and sore throat; they were treated and sent home for 24 h. The family members of the 14 GAS-positive HCWs were also screened; 7 of these family members were also positive for GAS. All persons who had cultures positive for GAS were treated with oral penicillin (500 mg 4 times per day for 10 days; doses for children were calculated according to body weight) [7]. Oral azithromycin was offered as an alternative for individuals allergic to penicillin.
Interviews with family members of the index patient and the respiratory therapist were conducted. Results of throat cultures were negative for these contacts.
Results made available on 1 March 2005 indicated identical PFGE patterns in isolates from the index patient and respiratory therapist (figure 1). Their isolates were susceptible to clindamycin and penicillin.
Pulsed-field gel electrophoresis of Streptococcus pyogenes isolated from blood samples from the index patient and an ill health care worker (courtesy of the Arizona Department of Health Services microbiology services).
The 2 case patients were infected with an identical clone, emm 1.6. The emm 1.6 clone varies by a single nucleotide substitution within the type-determining region from the common emm 1.0, which accounts for 99% of type emm 1s. The emm 1.6 marker is quite stable (B. Beall, personal communication).
When an infected patient has been documented to have transmitted GAS to an HCW, the disease has typically consisted of a mild to moderate illness, and rapid recovery occurs [8]. We report the apparent transmission of GAS from an index patient to a HCW that resulted in fulminant pneumonia and streptococcal toxic shock syndrome. We believe that the significance of this report is related to transmission occurring after the index patient had received appropriate antimicrobial therapy for >48 h and when the index patient had been in contact isolation. Because the index patient received mechanical ventilation with a closed system and had no clinical pneumonia, we surmise that transmission occurred through aerosols generated through the large open wounds of the index patient and the copious secretions produced around the time of the debridement surgery. The burden of organisms from the index patient may have led to incomplete killing at the wound site despite antimicrobial therapy, with prolonged shedding into the environment. There may be unknown host factors to account for the illness experienced by the respiratory therapist and not other HCWs with direct and prolonged contact with the index patient.
In 1990, a 30-year-old firefighter developed a GAS hand wound infection with streptococcal toxic shock syndrome after trying to resuscitate a 3-year-old child who died of overwhelming GAS disease. The firefighter was treated with penicillin G (2 million U IV every 4 h) and was discharged from the hospital. He was readmitted to the hospital 9 days after discharge (1 day after completion of antibiotic therapy) with pain, swelling, and erythema of the right hand. He was treated with penicillin G (2 million U IV 5 times per day) for 7 days and experienced no additional recurrence [9].
In a study by Lannigan et al. [10], a single patient transmitted GAS to 24 HCWs. Pharyngitis developed <4 days after exposure to the source patient who received IV vancomycin, clindamycin, and piperacillin-tazobactam. Intravenous immunoglobulin was given twice. Transmission apparently occurred after 48 h of antimicrobial therapy.
Chandler et al. [8] sighted current changes to their policy. For surgical debridement for patients with severe cutaneous infection, they recommend the use of N-95 masks for HCWs who might be intraoperatively exposed.
There are several limitations that may preclude making a definite link between the index patient and the respiratory therapist. Our contact investigation was performed after the PFGE results were available, 20 days after the admission of the index patient. The index patient's wound cultures were not performed postoperatively on a daily basis to document good tissue penetrations. However, standard maximum dosing for patients with renal insufficiency were used.
Because of the aforementioned studies and the fact that the respiratory therapist appeared to acquire invasive GAS disease from a patient who he had treated, we changed our policy to include contact and droplet precautions for all patients presenting with streptococcal toxic shock syndrome and necrotizing fasciitis until 48 h of antibiotic treatment was been received and until serial clinical specimens test negative for GAS.
We thank Shelly Kontz, for her help in this project; the Arizona Department of Health Services microbiology service; and Chris A. Van Beneden and Bernard Beall, for encouragement, assistance with genetic studies, and recommendations.
Potential conflicts of interest. All authors: no conflicts.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy