277
Clusters of Invasive Group A Streptococcal Infections in Family, Hospital, and Nursing Home Settings Benjamin Schwartz, John A. Elliott, Jay C. Butler, Paul A. Simon, Bonnie L. Jameson, Gregory E. Welch, and Richard R. Facklam
From the Respiratory Diseases Branch. Division ofBacterial and Mycotic Diseases. National Center for lnfectious Diseases. and the Division of Field Epidemiology. Centers for Disease Control. Atlanta. Georgia; the Bureau ofPublic Health. Wisconsin Department of Health and Social Services. Madison. Wisconsin; the Division of Disease Control and Environmental Epidemiology. Colorado Department of Health. Denver. Colorado; and the Division ofPublic Health and Communicable Disease Control. South Dakota Department of Health. Pierre, South Dakota
The ability of group A streptococci to spread to close contacts of infected patients has long been recognized. In families with a child who has pharyngitis, one or more secondary infections will commonly occur [1]. Although secondary cases of pharyngitis, except in the setting of rheumatic fever, are associated with minimal morbidity, the recent resurgence of severe, invasive group A streptococcal infections [2, 3] increases the potential for clusters of patients with invasive disease. Brief reports of a family cluster [4], nursing home clusters [5], and the spread of infection from a patient to a health care worker [6] have been published. In addition, spread of group A streptococcal infection from colonized health care personnel to patients, especially in surgical or gynecologic settings, is well recognized [7]. In this report we describe in detail clusters ofinvasive infections that occurred in a family, hospital, and nursing home, characterize the epidemiology of all clusters reported to the Centers for Disease
Received 26 December 1991: revised 12 March 1992. Reprints or correspondence: Dr. Benjamin Schwartz. Respiratory Diseases Branch. Mailstop C-09. Centers for Disease Control. 1600 Clifton Road N.E .. Atlanta. Georgia 30333. Clinical Infectious Diseases
1992;15:277-84
This article is in the public domain.
Control (CDC; Atlanta) since December 1988. and discuss their significance with regard to public health.
Reports of Clusters Family cluster. On 8 November 1990 a 25-year-old student was seen at a university infirmary for evaluation of a sore throat. He was afebrile and had nonexudative pharyngitis and cervical lymphadenopathy. A throat culture was positive for group A streptococci, and he received a 10-day course of oral penicillin. Between 21 and 25 November he returned home to visit his parents, during which time he had a sore throat. Two days after his return to the university, he again presented to the infirmary and had nonexudative pharyngitis without adenopathy. A repeated throat culture was positive for group A streptococci, and he completed a second IO-day course of penicillin. On 2 December, one week after the student's return to school, his father experienced the onset of perianal burning and redness. Four days later he noted a pustular lesion on his right index finger, and the following day he noted one on his left hand. Both hand lesions ruptured, resulting in drainage of yellow serous fluid. Although the perianal discomfort resolved, the hand lesions persisted through the following week.
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 21, 2015
The spread of group A streptococcal infection to close contacts of infected persons is well recognized. With the resurgence of invasive group A streptococcal infections, there is an increased potential for clusters of patients with invasive disease. We reviewed data collected since December 1988 at the Centers for Disease Control (Atlanta) to identify clusters of infection in which one or more patients had invasive disease. Twelve family clusters were identified. Infection in index cases included the toxic shock-like syndrome and septicemia. Infection in family contacts included invasive infections, pharyngitis, or asymptomatic carriage. Most invasive disease occurred in adults, while the majority of noninvasive infections were in children. Five nosocomial clusters with spread of infection from patients to hospital personnel were documented. All index patients had the toxic shock-like syndrome; secondary infections included the toxic shock-like syndrome, pneumonia, bullous cellulitis, lymphangitis, and pharyngitis. Clusters of invasive infections also were identified in five nursing homes. Pneumonia, cutaneous infections, and the toxic shock-like syndrome occurred most commonly. Clustering by nursing home unit occurred in three outbreaks. In hospitals and nursing homes, improved infection control will likely decrease secondary spread; in families, spread of disease may be prevented by identifying and treating those harboring the organism or by chemoprophylaxis. Studies that characterize the rate of secondary infection are needed before definitive recommendations can be made.
278
Schwartz et al.
developed a severe sore throat and fever. A throat swab culture was positive for group A Streptococcus. She was treated with oral penicillin, and the symptoms resolved. A second nurse who cared for the patient during two shifts on 8 October began to feel ill 3 days later, and on 13 October she developed a sore throat and fever. A throat swab was cultured on 15 October and was positive for group A Streptococcus; she also was treated with oral penicillin, and the symptoms resolved. On 12 October a 77-year-old man who had been in ICU bed 10 from 5 to 9 October following a myocardial infarction reported having a sore throat and cough. Two days later, a chest roentgenogram showed a right lower-lobe pneumonia. A throat swab culture from 13 October and sputum cultures from the following 2 days were positive for group A streptococci. He was transferred to the ICU on 15 October and required assisted ventilation. His condition improved with administration of intravenous antibiotics, and he was extubated on 21 October. Between 7 and 9 October this patient was situated two beds away from the index patient. Although neither of the nurses who became ill cared for him, brief cross-coverage of patients during breaks was common. On 13 October an 82-year-old man was admitted to ICU bed 11 following a hypoxic episode and seizure that occurred during bronchoscopy for evaluation of interstitial edema of unknown etiology. He was transferred out of the ICU the following day but became febrile, and therapy with ticarcillin/clavulanate was begun. Group A streptococci were isolated from a culture of sputum obtained on 15 October. A chest roentgenogram the following day showed left upperand lower-lobe infiltrates. His condition improved with antibiotic therapy. Records show that the second nurse described above cared for this patient in the ICU during the day and evening shifts on 13 October, the same day that she reported onset of a sore throat and fever. Following identification of this cluster on 15 October, respiratory isolation of infected patients was instituted. Throat swabs of all ICU patients and nurses were cultured; all were negative. No other nosocomial group A streptococcal infections were subsequently identified. Group A streptococcal isolates from this cluster were sent to the CDC for typing. All were M type 1, T type 1. All produced pyrogenic exotoxins A and B. Nursing home cluster. Nursing Home A is a 61-bed skilled nursing facility. Residents are housed in three wings, including a locked 14-bed Alzheimer's Disease Unit (ADU), which opened in August 1990. In December 12 patients ranging in age from 44 to 96 years resided in the unit. Nursing care was provided by six staff members who worked solely in the ADU and a similar number who worked throughout the facility. From the time the unit opened through 26 December, no group A streptococcal infection was diagnosed for any resident, and no patient had been
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 21, 2015
On 10 December the student's mother noted the onset ofa clear vaginal discharge, 2-3 days after last having had sexual intercourse with her husband. On 12 December she was seen by a gynecologist who prescribed a vaginal suppository for suspected vaginal candidiasis. Within 3 hours ofinserting the first suppository, she vomited and, suspecting a reaction to the the medication, used a vaginal douche. Over the next 24 hours, she continued to vomit and developed diarrhea, fever, and chills that persisted for 4 days. On 18 December she saw another physician and was admitted to the hospital. On admission, her temperature was 102.8°F; blood pressure, 120/69 mm Hg; and pulse rate, 104. Pelvic examination revealed yellow drainage from the cervical os and mild adnexal tenderness. A chest roentgenogram showed small, bilateral pleural effusions without infiltrates. Intravenous clindamycin and gentamicin were administered. On the second hospital day, blood cultures and culture offluid from the cervical os yielded group A streptococci, and therapy was changed to intravenous penicillin. Her condition improved clinically, and she was discharged after receiving a 10-day course of intravenous therapy. After the isolation of group A streptococci from the mother's blood, this organism was isolated from the lesions on the father's hands and, again, from the student's throat. The father was treated with intramuscular benzathine penicillin and the lesions resolved within I week. The student received a 10-day course of oral cephalexin, and a follow-up culture was negative. The group A streptococcal isolates from the mother and father were serotyped by the Bacterial Reference Laboratory at the CDC; both were M type 3, T type 3/13. Nosocomial cluster. Hospital A is a 565-bed community teaching hospital with a 19-bed intensive care unit (lCU). Eleven of the ICU beds are arranged in a semicircle and are spaced 11 feet apart and separated by curtains. On 7 October 1989 an 84-year-old man was admitted to ICU bed 8 with a I-day history of rapidly progressive facial swelling involving the lips, cheeks, and periorbital area. His temperature on admission was 100.5°F. A chest roentengogram showed bibasilar infiltrates, and therapy with intravenous erythromycin and clindamycin was begun. During the next 24 hours, he developed hypotension, renal failure, and peripheral edema. Cultures of two blood samples drawn on admission were positive for group A Streptococcus. The patient's cellulitis and pneumonia gradually abated, and on 13 October he was transferred out of the ICU; however, his renal function continued to deteriorate, and he died on I November. The nurse who admitted this patient on 7 October cared for him for I hour and had no other documented contact. During this time, she recalled that he coughed frequently, especially when she obtained a specimen for a throat culture. Aside from universal precautions, no additional isolation measures were in effect at the time. On 9 October this nurse
CIO 1992;15 (August)
CIO 1992; 15 (August)
Invasive Group A Streptococcal Infections
testing for pyrogenic exotoxin production. All isolates except that from the second patient were M type 3, T type 3/13. The isolate from the second patient was an M nontypable, T nontypable strain. The M type 3 isolates produced both pyrogenic exotoxins A and B. Whole-cell immunoblotting confirmed the difference between the second patient's strain and the other isolates.
Methods The Bacterial Reference Laboratory at the CDC has served as a reference center for group A streptococci since 1949. Isolates are sent for serotyping at the discretion ofphysicians and state health departments. All isolates are accompanied by a brief form on which the patient's age and clinical diagnosis and the site of isolation of the organism are noted. A comment section may provide additional clinical and epidemiological information. In April 1990 the CDC attempted to enhance surveillance for invasive group A streptococcal infection by encouraging state health departments to use a standard form for reporting detailed clinical information about cases. This form has accompanied some of the isolates sent for typing. We identified potential clusters of infection from isolates received since December 1988 on the basis of information included on the laboratory or clinical report forms or from discussions with local physicians or health officers. Invasive infections were defined by the isolation of group A streptococci from a normally sterile site or by the isolation of the organisms from a non sterile site in the presence of systemic symptoms. Patients were designated as having the toxic shock-like syndrome on the basis of the physicians' reports. Since in many ofthe clusters the temporal sequence ofacquisition of group A streptococci could not be determined, we have defined the index patient as the first patient in each cluster reported to the CDC. Clusters were confirmed by documenting identical M and T types among epidemiologically linked isolates. Typing was performed using standard methods [8] with antisera produced at the CDC and tested with reference strains. For some clusters additional laboratory studies were performed to confirm that isolates were identical: production of pyrogenic exotoxins A and B was determined by the western immunoblot technique with use of rabbit polyclonal antisera, and wholecell protein patterns were determined by SDS-PAGE [9]. The significance of differences was tested for dichotomous variables using the X2 test and for continuous variables using the Kruskal-Wallis test.
Results Family clusters. Since December 1988 12 household clusters, including the one described above, have been reported to the CDC (table I). Two patients with sepsis also
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 21, 2015
hospitalized with an infectious condition of unknown etiology. On 25 December a 90-year-old man was noted by the nursing staff to be lethargic. His temperature was 101.4"F. No other symptoms were noted. The following day he remained febrile and developed a macular erythematous rash on his right leg. He was transferred to the hospital on 27 December and died the same day. Cultures of blood obtained on admission were positive for Streptococcus pyogenes. The second case occurred 9 days after the onset of illness in the index patient. On 3 January an 80-year-old woman reported having fallen. Although no traumatic injury was noted, she became febrile (temperature to 101°F) II hours later. The next day she developed signs ofmild upper respiratory tract illness, including a nonproductive cough. These signs persisted, and on 6 January she was transferred to the hospital. On admission she was febrile (temperature to 105.5°F), and bilateral infiltrates were seen on a chest roentgenogram; a blood culture yielded group A streptococci. Her illness progressed rapidly, resulting in multiple organ-system failure, and she died later that day. Four days later a 96-year-old woman was found to have a low-grade fever and serous drainage from shallow ulcerations on her leg. Blood, throat, and discharge cultures were positive for group A Streptococcus. Therapy was started with oral cephalexin, and she recovered uneventfully without requiring hospitalization. The fourth patient, an 82-year-old man, developed symptoms of upper respiratory tract infection on 12 January and became febrile (temperature to 102.8°F) on 14 January. He was transferred to the hospital, where a chest roentgenogram indicated a right upper-lobe pneumonia. A throat culture prior to transfer and a culture of blood obtained on admission yielded group A Streptococcus. He recovered from the infection with administration of antimicrobial therapy. On 14 January throat swabs were obtained for culture from all residents of the ADU and all staff members who cared for them. One asymptomatic resident and one staff person were identified as carriers. Antimicrobial agents were administered to the individuals for whom cultures were positive and were given prophylactically to several other residents of the unit. No further cases of streptococcal disease have been diagnosed. Nursing staff members at the facility were questioned to identify potential risk factors for the development ofgroup A streptococcal infection among residents. All fiveinfected residents had separate rooms, and each had an uninfected roommate. All were ambulatory, and although close contact between residents was common, the infected patients did not have closer contact with one another than with other residents. No resident of the unit was immunosuppressed and none was taking steroids. All group A streptococcal isolates were sent to the Bacterial Reference Laboratory at the CDC for serotyping and
279
Schwartz et at.
280
CID 1992; 15 (August)
Table 1. Summary of data on family clusters of invasive group A streptococcal infection reported to the CDC from December 1988 to March 1991. Cases of secondary infection Index patient
Cluster I 2
3 4
9
to II 12
Age (y)/sex Illness
Illinois California California Colorado New Hampshire Missouri Colorado Ohio
171M 821M 42/F
Michigan California Wisconsin Michigan
65/M 29/F
NOTE.
35/M
291M 32/F 67/F 39/F
55/F 9fM
Culture site
TSLS Blood TSLS Blood Sepsis Blood Sepsis Blood Sepsis Blood TSLS Blood Sepsis Blood TSLS Blood, sputum Sepsis Blood TSLS Blood Sepsis Blood TSLS Throat
TSLS = toxic shock-like syndrome; NT
Age Outcome (y)/sex
Illness
Died Died Died Died Survived Survived Survived Survived
121M Sepsis 79/F TSLS
Survived Survived Survived ?
85/F I d/F 57/M
Culture site Blood Tissue
Noninvasive" Number Children Outcome infected included Survived Died
2
Yes
2
3 1 4
Yes Yes No Yes No Yes
Sputum CSF Throat Pleural fluid
Survived Died Survived ?
3
Yes
M~3, T~3 M~L T~I
2 I
Pneumonia Meningitis Cellulitis 121M TSLS
Serotype
No
M~I,
T-I
M~I, T~I
M-NT, T-5/27/44 M-NT, T-4/28 M-l, T~I M-NT, T-NT M-12, T-12 M-l, r-t M-NT, T-NT M-5, T-5/27/44
nontypable; ? = not reported.
* Pharyngitis or asymptomatic pharyngeal carriage.
had pneumonia and one had cellulitis. The median age of index patients was 35 years. Infection was fatal in four (36%) of II index patients for whom the outcome of illness was reported. In each cluster, from one to four other household contacts were infected. Six of the clusters involved secondary cases of invasive infection (table I). Two (40%) of five infected contacts for whom the outcome was known died: an index patient's 79-year-old spouse who had the toxic shock-like syndrome and an infant who developed meningitis on the first day oflife after being born to a mother with the toxic shocklike syndrome. The median age offamily contacts with invasive infection was 57 years (range, 0-85 years). In four clusters one or more family contacts had symptomatic pharyngitis before onset of symptoms in the patient with invasive infection; however, only in the family cluster described above were cultures performed for this person or was the patient treated before the case of invasive infection occurred. In the remaining clusters, no symptoms were reported for culture-positive contacts. Secondary illness occurred between I and 14 days after the onset ofsymptoms in the index patient. Overall, eight (67%) of the 12 clusters involved persons (either index patients or family contacts) who were
Clusters of Invasive Group A Streptococcal Infections in Family, Hospital, and Nursing Home Settings Benjamin Schwartz, John A. Elliott, Jay C. Butler, Paul A. Simon, Bonnie L. Jameson, Gregory E. Welch, and Richard R. Facklam
From the Respiratory Diseases Branch. Division ofBacterial and Mycotic Diseases. National Center for lnfectious Diseases. and the Division of Field Epidemiology. Centers for Disease Control. Atlanta. Georgia; the Bureau ofPublic Health. Wisconsin Department of Health and Social Services. Madison. Wisconsin; the Division of Disease Control and Environmental Epidemiology. Colorado Department of Health. Denver. Colorado; and the Division ofPublic Health and Communicable Disease Control. South Dakota Department of Health. Pierre, South Dakota
The ability of group A streptococci to spread to close contacts of infected patients has long been recognized. In families with a child who has pharyngitis, one or more secondary infections will commonly occur [1]. Although secondary cases of pharyngitis, except in the setting of rheumatic fever, are associated with minimal morbidity, the recent resurgence of severe, invasive group A streptococcal infections [2, 3] increases the potential for clusters of patients with invasive disease. Brief reports of a family cluster [4], nursing home clusters [5], and the spread of infection from a patient to a health care worker [6] have been published. In addition, spread of group A streptococcal infection from colonized health care personnel to patients, especially in surgical or gynecologic settings, is well recognized [7]. In this report we describe in detail clusters ofinvasive infections that occurred in a family, hospital, and nursing home, characterize the epidemiology of all clusters reported to the Centers for Disease
Received 26 December 1991: revised 12 March 1992. Reprints or correspondence: Dr. Benjamin Schwartz. Respiratory Diseases Branch. Mailstop C-09. Centers for Disease Control. 1600 Clifton Road N.E .. Atlanta. Georgia 30333. Clinical Infectious Diseases
1992;15:277-84
This article is in the public domain.
Control (CDC; Atlanta) since December 1988. and discuss their significance with regard to public health.
Reports of Clusters Family cluster. On 8 November 1990 a 25-year-old student was seen at a university infirmary for evaluation of a sore throat. He was afebrile and had nonexudative pharyngitis and cervical lymphadenopathy. A throat culture was positive for group A streptococci, and he received a 10-day course of oral penicillin. Between 21 and 25 November he returned home to visit his parents, during which time he had a sore throat. Two days after his return to the university, he again presented to the infirmary and had nonexudative pharyngitis without adenopathy. A repeated throat culture was positive for group A streptococci, and he completed a second IO-day course of penicillin. On 2 December, one week after the student's return to school, his father experienced the onset of perianal burning and redness. Four days later he noted a pustular lesion on his right index finger, and the following day he noted one on his left hand. Both hand lesions ruptured, resulting in drainage of yellow serous fluid. Although the perianal discomfort resolved, the hand lesions persisted through the following week.
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 21, 2015
The spread of group A streptococcal infection to close contacts of infected persons is well recognized. With the resurgence of invasive group A streptococcal infections, there is an increased potential for clusters of patients with invasive disease. We reviewed data collected since December 1988 at the Centers for Disease Control (Atlanta) to identify clusters of infection in which one or more patients had invasive disease. Twelve family clusters were identified. Infection in index cases included the toxic shock-like syndrome and septicemia. Infection in family contacts included invasive infections, pharyngitis, or asymptomatic carriage. Most invasive disease occurred in adults, while the majority of noninvasive infections were in children. Five nosocomial clusters with spread of infection from patients to hospital personnel were documented. All index patients had the toxic shock-like syndrome; secondary infections included the toxic shock-like syndrome, pneumonia, bullous cellulitis, lymphangitis, and pharyngitis. Clusters of invasive infections also were identified in five nursing homes. Pneumonia, cutaneous infections, and the toxic shock-like syndrome occurred most commonly. Clustering by nursing home unit occurred in three outbreaks. In hospitals and nursing homes, improved infection control will likely decrease secondary spread; in families, spread of disease may be prevented by identifying and treating those harboring the organism or by chemoprophylaxis. Studies that characterize the rate of secondary infection are needed before definitive recommendations can be made.
278
Schwartz et al.
developed a severe sore throat and fever. A throat swab culture was positive for group A Streptococcus. She was treated with oral penicillin, and the symptoms resolved. A second nurse who cared for the patient during two shifts on 8 October began to feel ill 3 days later, and on 13 October she developed a sore throat and fever. A throat swab was cultured on 15 October and was positive for group A Streptococcus; she also was treated with oral penicillin, and the symptoms resolved. On 12 October a 77-year-old man who had been in ICU bed 10 from 5 to 9 October following a myocardial infarction reported having a sore throat and cough. Two days later, a chest roentgenogram showed a right lower-lobe pneumonia. A throat swab culture from 13 October and sputum cultures from the following 2 days were positive for group A streptococci. He was transferred to the ICU on 15 October and required assisted ventilation. His condition improved with administration of intravenous antibiotics, and he was extubated on 21 October. Between 7 and 9 October this patient was situated two beds away from the index patient. Although neither of the nurses who became ill cared for him, brief cross-coverage of patients during breaks was common. On 13 October an 82-year-old man was admitted to ICU bed 11 following a hypoxic episode and seizure that occurred during bronchoscopy for evaluation of interstitial edema of unknown etiology. He was transferred out of the ICU the following day but became febrile, and therapy with ticarcillin/clavulanate was begun. Group A streptococci were isolated from a culture of sputum obtained on 15 October. A chest roentgenogram the following day showed left upperand lower-lobe infiltrates. His condition improved with antibiotic therapy. Records show that the second nurse described above cared for this patient in the ICU during the day and evening shifts on 13 October, the same day that she reported onset of a sore throat and fever. Following identification of this cluster on 15 October, respiratory isolation of infected patients was instituted. Throat swabs of all ICU patients and nurses were cultured; all were negative. No other nosocomial group A streptococcal infections were subsequently identified. Group A streptococcal isolates from this cluster were sent to the CDC for typing. All were M type 1, T type 1. All produced pyrogenic exotoxins A and B. Nursing home cluster. Nursing Home A is a 61-bed skilled nursing facility. Residents are housed in three wings, including a locked 14-bed Alzheimer's Disease Unit (ADU), which opened in August 1990. In December 12 patients ranging in age from 44 to 96 years resided in the unit. Nursing care was provided by six staff members who worked solely in the ADU and a similar number who worked throughout the facility. From the time the unit opened through 26 December, no group A streptococcal infection was diagnosed for any resident, and no patient had been
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 21, 2015
On 10 December the student's mother noted the onset ofa clear vaginal discharge, 2-3 days after last having had sexual intercourse with her husband. On 12 December she was seen by a gynecologist who prescribed a vaginal suppository for suspected vaginal candidiasis. Within 3 hours ofinserting the first suppository, she vomited and, suspecting a reaction to the the medication, used a vaginal douche. Over the next 24 hours, she continued to vomit and developed diarrhea, fever, and chills that persisted for 4 days. On 18 December she saw another physician and was admitted to the hospital. On admission, her temperature was 102.8°F; blood pressure, 120/69 mm Hg; and pulse rate, 104. Pelvic examination revealed yellow drainage from the cervical os and mild adnexal tenderness. A chest roentgenogram showed small, bilateral pleural effusions without infiltrates. Intravenous clindamycin and gentamicin were administered. On the second hospital day, blood cultures and culture offluid from the cervical os yielded group A streptococci, and therapy was changed to intravenous penicillin. Her condition improved clinically, and she was discharged after receiving a 10-day course of intravenous therapy. After the isolation of group A streptococci from the mother's blood, this organism was isolated from the lesions on the father's hands and, again, from the student's throat. The father was treated with intramuscular benzathine penicillin and the lesions resolved within I week. The student received a 10-day course of oral cephalexin, and a follow-up culture was negative. The group A streptococcal isolates from the mother and father were serotyped by the Bacterial Reference Laboratory at the CDC; both were M type 3, T type 3/13. Nosocomial cluster. Hospital A is a 565-bed community teaching hospital with a 19-bed intensive care unit (lCU). Eleven of the ICU beds are arranged in a semicircle and are spaced 11 feet apart and separated by curtains. On 7 October 1989 an 84-year-old man was admitted to ICU bed 8 with a I-day history of rapidly progressive facial swelling involving the lips, cheeks, and periorbital area. His temperature on admission was 100.5°F. A chest roentengogram showed bibasilar infiltrates, and therapy with intravenous erythromycin and clindamycin was begun. During the next 24 hours, he developed hypotension, renal failure, and peripheral edema. Cultures of two blood samples drawn on admission were positive for group A Streptococcus. The patient's cellulitis and pneumonia gradually abated, and on 13 October he was transferred out of the ICU; however, his renal function continued to deteriorate, and he died on I November. The nurse who admitted this patient on 7 October cared for him for I hour and had no other documented contact. During this time, she recalled that he coughed frequently, especially when she obtained a specimen for a throat culture. Aside from universal precautions, no additional isolation measures were in effect at the time. On 9 October this nurse
CIO 1992;15 (August)
CIO 1992; 15 (August)
Invasive Group A Streptococcal Infections
testing for pyrogenic exotoxin production. All isolates except that from the second patient were M type 3, T type 3/13. The isolate from the second patient was an M nontypable, T nontypable strain. The M type 3 isolates produced both pyrogenic exotoxins A and B. Whole-cell immunoblotting confirmed the difference between the second patient's strain and the other isolates.
Methods The Bacterial Reference Laboratory at the CDC has served as a reference center for group A streptococci since 1949. Isolates are sent for serotyping at the discretion ofphysicians and state health departments. All isolates are accompanied by a brief form on which the patient's age and clinical diagnosis and the site of isolation of the organism are noted. A comment section may provide additional clinical and epidemiological information. In April 1990 the CDC attempted to enhance surveillance for invasive group A streptococcal infection by encouraging state health departments to use a standard form for reporting detailed clinical information about cases. This form has accompanied some of the isolates sent for typing. We identified potential clusters of infection from isolates received since December 1988 on the basis of information included on the laboratory or clinical report forms or from discussions with local physicians or health officers. Invasive infections were defined by the isolation of group A streptococci from a normally sterile site or by the isolation of the organisms from a non sterile site in the presence of systemic symptoms. Patients were designated as having the toxic shock-like syndrome on the basis of the physicians' reports. Since in many ofthe clusters the temporal sequence ofacquisition of group A streptococci could not be determined, we have defined the index patient as the first patient in each cluster reported to the CDC. Clusters were confirmed by documenting identical M and T types among epidemiologically linked isolates. Typing was performed using standard methods [8] with antisera produced at the CDC and tested with reference strains. For some clusters additional laboratory studies were performed to confirm that isolates were identical: production of pyrogenic exotoxins A and B was determined by the western immunoblot technique with use of rabbit polyclonal antisera, and wholecell protein patterns were determined by SDS-PAGE [9]. The significance of differences was tested for dichotomous variables using the X2 test and for continuous variables using the Kruskal-Wallis test.
Results Family clusters. Since December 1988 12 household clusters, including the one described above, have been reported to the CDC (table I). Two patients with sepsis also
Downloaded from http://cid.oxfordjournals.org/ at University of Sussex on August 21, 2015
hospitalized with an infectious condition of unknown etiology. On 25 December a 90-year-old man was noted by the nursing staff to be lethargic. His temperature was 101.4"F. No other symptoms were noted. The following day he remained febrile and developed a macular erythematous rash on his right leg. He was transferred to the hospital on 27 December and died the same day. Cultures of blood obtained on admission were positive for Streptococcus pyogenes. The second case occurred 9 days after the onset of illness in the index patient. On 3 January an 80-year-old woman reported having fallen. Although no traumatic injury was noted, she became febrile (temperature to 101°F) II hours later. The next day she developed signs ofmild upper respiratory tract illness, including a nonproductive cough. These signs persisted, and on 6 January she was transferred to the hospital. On admission she was febrile (temperature to 105.5°F), and bilateral infiltrates were seen on a chest roentgenogram; a blood culture yielded group A streptococci. Her illness progressed rapidly, resulting in multiple organ-system failure, and she died later that day. Four days later a 96-year-old woman was found to have a low-grade fever and serous drainage from shallow ulcerations on her leg. Blood, throat, and discharge cultures were positive for group A Streptococcus. Therapy was started with oral cephalexin, and she recovered uneventfully without requiring hospitalization. The fourth patient, an 82-year-old man, developed symptoms of upper respiratory tract infection on 12 January and became febrile (temperature to 102.8°F) on 14 January. He was transferred to the hospital, where a chest roentgenogram indicated a right upper-lobe pneumonia. A throat culture prior to transfer and a culture of blood obtained on admission yielded group A Streptococcus. He recovered from the infection with administration of antimicrobial therapy. On 14 January throat swabs were obtained for culture from all residents of the ADU and all staff members who cared for them. One asymptomatic resident and one staff person were identified as carriers. Antimicrobial agents were administered to the individuals for whom cultures were positive and were given prophylactically to several other residents of the unit. No further cases of streptococcal disease have been diagnosed. Nursing staff members at the facility were questioned to identify potential risk factors for the development ofgroup A streptococcal infection among residents. All fiveinfected residents had separate rooms, and each had an uninfected roommate. All were ambulatory, and although close contact between residents was common, the infected patients did not have closer contact with one another than with other residents. No resident of the unit was immunosuppressed and none was taking steroids. All group A streptococcal isolates were sent to the Bacterial Reference Laboratory at the CDC for serotyping and
279
Schwartz et at.
280
CID 1992; 15 (August)
Table 1. Summary of data on family clusters of invasive group A streptococcal infection reported to the CDC from December 1988 to March 1991. Cases of secondary infection Index patient
Cluster I 2
3 4
9
to II 12
Age (y)/sex Illness
Illinois California California Colorado New Hampshire Missouri Colorado Ohio
171M 821M 42/F
Michigan California Wisconsin Michigan
65/M 29/F
NOTE.
35/M
291M 32/F 67/F 39/F
55/F 9fM
Culture site
TSLS Blood TSLS Blood Sepsis Blood Sepsis Blood Sepsis Blood TSLS Blood Sepsis Blood TSLS Blood, sputum Sepsis Blood TSLS Blood Sepsis Blood TSLS Throat
TSLS = toxic shock-like syndrome; NT
Age Outcome (y)/sex
Illness
Died Died Died Died Survived Survived Survived Survived
121M Sepsis 79/F TSLS
Survived Survived Survived ?
85/F I d/F 57/M
Culture site Blood Tissue
Noninvasive" Number Children Outcome infected included Survived Died
2
Yes
2
3 1 4
Yes Yes No Yes No Yes
Sputum CSF Throat Pleural fluid
Survived Died Survived ?
3
Yes
M~3, T~3 M~L T~I
2 I
Pneumonia Meningitis Cellulitis 121M TSLS
Serotype
No
M~I,
T-I
M~I, T~I
M-NT, T-5/27/44 M-NT, T-4/28 M-l, T~I M-NT, T-NT M-12, T-12 M-l, r-t M-NT, T-NT M-5, T-5/27/44
nontypable; ? = not reported.
* Pharyngitis or asymptomatic pharyngeal carriage.
had pneumonia and one had cellulitis. The median age of index patients was 35 years. Infection was fatal in four (36%) of II index patients for whom the outcome of illness was reported. In each cluster, from one to four other household contacts were infected. Six of the clusters involved secondary cases of invasive infection (table I). Two (40%) of five infected contacts for whom the outcome was known died: an index patient's 79-year-old spouse who had the toxic shock-like syndrome and an infant who developed meningitis on the first day oflife after being born to a mother with the toxic shocklike syndrome. The median age offamily contacts with invasive infection was 57 years (range, 0-85 years). In four clusters one or more family contacts had symptomatic pharyngitis before onset of symptoms in the patient with invasive infection; however, only in the family cluster described above were cultures performed for this person or was the patient treated before the case of invasive infection occurred. In the remaining clusters, no symptoms were reported for culture-positive contacts. Secondary illness occurred between I and 14 days after the onset ofsymptoms in the index patient. Overall, eight (67%) of the 12 clusters involved persons (either index patients or family contacts) who were
