Household transmission of Streptococcus pneumoniae, Alberta, Canada.

Proven or presumptive multidrug-resistant Streptococcus pneumoniae pneumonia was diagnosed simultaneously in three married couples in Alberta, Canada. The pair of isolates from each couple had identical antibiotic resistance profiles, serotypes, and pulsed-field gel electrophoresis patterns. One or more of these cases could have been prevented by S. pneumoniae vaccine.

In December 1996 and January 1997, three married couples with multidrug-resistant S. pneumoniae (MDRSP) were admitted to Foothills Medical Centre in Calgary. The couples were not admitted on the same day. None of the couples lived with children, although couple C had daily contact with children. All patients received appropriate antibiotic therapy after their culture and antibiotic sensitivity results were known. We reviewed each patients health record (Table) and were able to contact two of the three couples for further information.
S. pneumoniae were identified by standard methods. MICs were determined by E-Test (AB Biodisk, Solna, Sweden) and classified as susceptible (S), intermediate resistant (I), or fully resistant (R) to each antibiotic, according to National Committee for Clinical Laboratory Standards guidelines (12). Serotyping of S. pneumoniae was performed by the Quellung reaction technique at the National Centre for Streptococcus, Edmonton. Electrophoretic fingerprinting of S. pneumoniae was performed by pulsed-field gel electrophoresis (PFGE) of DNA digested with Sma1 (BRL, Gaithersburg, MD). The PFGE patterns were classified as indistinguishable, related, or different according to criteria suggested by Tenover (13).
The diagnosis of S. pneumoniae pneumonia in couple A was confirmed by positive blood cultures, chest X-ray lobar pneumonia, and disease-compatible clinical findings. Patient 1 in couple A was a health records clerk at Foothills Medical Centre. Her illness was complicated soon after admission by empyema, which was drained; the fluid was S. pneumoniae-negative. Vertebral osteomyelitis was suspected from clinical evidence 18 days after admission and was confirmed by bone scan; no diagnostic culture was obtained. Osteomyelitis in this patient was presumably caused by S. pneumoniae. The initial 7-day course of cefuroxime (to which S. pneumoniae was resistant) may not have cleared the infection and thus allowed secondary seeding to bone.

Dispatches
Couple B (who could not be reached for further information) had had recent visitors from Texas (one a hospital worker) with upper respiratory tract infections. S. pneumoniae pneumonia was presumptively diagnosed in this couple on the basis of symptoms, signs, and chest X-rays compatible with the diagnosis of pneumonia, as well as sputum samples, which had gram-positive lancet-shaped cocci identified on Gram stain and grew S. pneumoniae. From the sputum of patient 2 in couple B, gramnegative bacilli were identified on Gram stain; Haemophilus influenzae was also isolated. Thus, this patient may have been coinfected, or primarily infected, with H. influenzae. The patients blood cultures were negative; a blood culture was not performed on patient 1 in couple B.
Couple C was admitted with severe burns and inhalation injuries after the stove in their two-room trailer exploded. They had had recurrent sinusitis and other respiratory infections in the previous year since moving to their trailer, which had poor air circulation. Patient 1 of this couple was taking antibiotics at the time of admission, and patient 2 had recently completed a course of antibiotics. The diagnosis of pneumonia (patient 1 on day 3 of admission and patient 2 on day 2) was made on the basis of recent upper respiratory symptoms and fever, diminished breath sounds, crepitations, and disease-compatible chest X-ray findings (previous films had been normal), which made pneumonia more likely than noninfectious conditions such as acute lung syndrome. The presumptive diagnosis of S. pneumoniae as the etiologic agent in the case of patient 1, couple C, was made on the basis of the initial endotracheal tube aspirate, which had gram-positive lancetshaped cocci identified on Gram stain and grew S. pneumoniae. Only gram-positive lancetshaped cocci were identified from the initial bronchoalveolar lavage of patient 2 on Gram stain, and S. pneumoniae grew in much greater numbers than H. influenzae. Blood cultures, performed for couple C only after antibiotic therapy was started, were negative. Patient 2 died of septic shock 20 days after admission, with Candida albicans in his blood. The bronchopneumonia never resolved clinically, although S. pneumoniae was not isolated from any further cultures. Thus, S. pneumoniae may have been a contributing factor to, but not likely the direct cause, of this patients death.
The identical susceptibility patterns, serotypes, and PFGE patterns indicate that both partners in each couple were infected with the same multidrug-resistant S. pneumoniae strain. Couples A and B apparently had communityacquired pneumonia. Although couple C contracted pneumonia 48 to 72 hours after admission, each partner entered the hospital already infected with MDRSP; the infecting organisms were identical, and no other recognized cases of nosocomial MDRSP occurred at Foothills Medical Centre at the time of their admission (they were admitted 1 month before couple B, who were also infected with serotype 9V MDRSP). Couple A may have been exposed to MDRSP as a result of one partners work in a tertiary-care hospital; couple B as a result of one partners exposure to a health-care worker with respiratory symptoms. At the time of these cases, the prevalence of penicillin-nonsusceptible S. pneumoniae infections in Calgary was approximately 10% (A.P. Gibb, unpub. data).
None of these patients had received S. pneumoniae vaccine, yet each had one or more risk factors for infection (advanced age, exposure to young children, smoking, and chronic lung or heart disease). Couple C had a history of recent antibiotic use, the predominant risk factor for antibiotic-resistant infections.
In Canada, the S. pneumoniae vaccine is recommended for all persons ≥65 years old and persons ≥2 years with identified risk factors (14). Despite the vaccines reasonable effectiveness, its use has been very low in Canada until recently (fewer than 12 doses per 10,000 population distributed annually [15,16]). The vaccine has been provided free of charge to persons with medical indications, but not to healthy persons 65 years of age and older and not as part of a routine vaccination schedule (17). Some provinces (including Alberta, beginning in 1998) have begun to routinely provide the vaccine to all persons at risk. The current incidence of invasive S. pneumoniae infections in Calgary is 20 per 100,000 per year overall and 87 per 100,000 per year in those older than 64 years of age (J.D. Klein, unpub. data).
Outbreaks of S. pneumoniae disease occur in institutions with crowding, poor air quality, or increased host susceptibility (2,4,6). These factors may also exist within households (9,11). Couple C, for example, lived in a very crowded space with poor air circulation.

Dispatches
The rate at which secondary S. pneumoniae infections occur in household contacts of index patients with invasive disease is not known, but rare cases have been reported (7)(8)(9)(10)(11). Factors contributing to secondary infections include the likelihood of nasopharyngeal infection due to exposure to the index patient or a common source, susceptibility to the strain of the index infection, and likelihood that colonization will lead to disease rather than to development of asymptomatic immunity. Data on contemporaneous nasopharyngeal carriage of the outbreak strain by household contacts are limited. A recent study from Gambia found carriage in 8.5% of household contacts, compared with 21% in an older U.S. study (18,19). In healthy adults, the prevalence of circulating S. pneumoniae antibodies is low (4% to 34%, depending on the serotype); however, two thirds of adults have protective antibody within 1 month of colonization (20). Approximately 15% of children who acquire a new S. pneumoniae strain nasopharyngeally in a nonoutbreak setting acquire clinical disease (usually otitis media); this rate is unknown for adults (21). In contrast, during a recent nursinghome pneumonia outbreak, 23% of residents were infected with the S. pneumoniae outbreak strain, and 4% became ill (22). The median age of residents was 85 years; only 4% had received S. pneumoniae vaccine.
Increased use of S. pneumoniae vaccine may prevent MDRSP pneumonia within households and among persons living in crowded conditions.