Identification of clinical risk factors for nosocomial pneumococcal bacteremia

Hospital-onset adult invasive pneumococcal disease in Israel: Sicker patients, different pathogens

OBJECTIVES

Invasive pneumococcal disease (IPD) usually has its onset in the community (CO-IPD), but it can commence following hospitalization (HO-IPD). This study compared HO-IPD and CO-IPD cases during the implementation of the pneumococcal conjugate vaccine (PCV) program for children in Israel.

METHODS

This was a nationwide retrospective cohort study of adult (age >18 years) IPD patients covering the period from the implementation of the PCV7/13 program in 2009/2010 through 2015. HO-IPD and CO-IPD were defined as IPD with onset ≥4 and ≤2 days from admission, respectively. Patient characteristics, outcome measures, serotypes, and antimicrobial susceptibility were compared for the entire cohort, followed by a matched case-control analysis.

RESULTS

The study included 114 patients with HO-IPD and 2180 with CO-IPD. After matching HO-IPD to CO-IPD patients by age, sex, and comorbidities, the mortality rate and discharge to long-term care facility rate were significantly higher for HO-IPD patients than for CO-IPD patients (44.6% vs. 26.3% and 26.5% vs. 8.2%, respectively). HO-IPD isolates were less often covered by PCV13 (39.6% vs. 49.0%) and pneumococcal polysaccharide vaccine PPSV23 (56.6% vs. 71.3%) and more often resistant to penicillin (9.3% vs. 3.6%), ceftriaxone (3.8% vs. 0.75%), and levofloxacin (9.3% vs. 0.8%).

CONCLUSIONS

HO-IPD was associated with higher morbidity and mortality than CO-IPD and was more often caused by non-vaccine serotypes (primarily non-PCV13 types) and antibiotic-resistant strains.

Description of Streptococcus pneumoniae infections in burn patients

BACKGROUND

Longer survival in burn patients has resulted in more infectious complications, typically with Klebsiella pneumoniae, Pseudomonas aeruginosa, Acinetobacter baumannii, and Staphylococcus aureus. Although Streptococcus pneumoniae infections are common in the community and can cause nosocomial infections, the incidence and risk factors for pneumococcal infections in burn patients is unclear.

METHODS

We performed an electronic retrospective chart review to collect rates of and risk factors for S. pneumoniae infections in patients with thermal burns from March 2003 through June 2008.

RESULTS

Of the 1838 patients admitted to the burn center, 10 were infected (0.54% incidence). Patients presented with pneumonia (seven patients, 0.38% incidence) and bacteremia (three patients, 0.16% incidence) within a week of initial burn (median 1 day, range 0-8), often in the setting of bacterial co-infection (five patients). This group was mainly young males with median 28.8% total body surface area burns; 60% had concomitant inhalational injury. Most did not have traditional risk factors for pneumococcal infection but had objective signs of infection at time of positive culture and were treated with appropriate antibiotics. Two patients in this series died, although no mortality was attributed to S. pneumoniae.

CONCLUSIONS

Pneumococcal disease is not common in burn patients and generally occurs early on in hospitalization after burn making it more likely to be a community-acquired pathogen rather than nosocomial in the burn population. It should be considered in the setting of sepsis or new pulmonary infiltrates within a week after burn, but typical empiric antibiotics against the usual burn pathogens should be adequate to also treat for pneumococcal infection.

Association of serotypes of Streptococcus pneumoniae with disease severity and outcome in adults: an international study

BACKGROUND

The introduction of conjugate pneumococcal vaccination for children has reduced the burden of invasive disease due to pneumococcal conjugate vaccine (PCV) types (i.e., serotypes 9V, 14, 6B, 18C, 23F, 19F, and 4) in adults. As nonvaccine serotypes become predominant causes of invasive disease among adults, it is necessary to evaluate the disease severity and mortality associated with infection due to nonvaccine serotypes, compared with PCV serotypes, in adults.

METHODS

The association of pneumococcal serotype and host-related variables with disease severity and mortality was statistically examined (with multivariable analysis) in 796 prospectively enrolled, hospitalized adult patients with bacteremia due to Streptococcus pneumoniae.

RESULTS

In multivariate analyses of risk in patients with invasive pneumococcal disease, older age (age, > or = 65 years; P = .004), underlying chronic disease (P = .025), immunosuppression (P = .035), and severity of disease (P < .001) were significantly associated with mortality; no association was found between nosocomial infection with invasive serotypes 1, 5, and 7 and mortality. The risk factors meningitis (P = .001), suppurative lung complications (P < or = .001), and preexisting lung disease (P = .051) were significantly associated with disease severity, independent of infecting serotype. No differences were seen in disease severity or associated mortality among patients infected with PCV serotypes, compared with patients infected with nonvaccine serotypes.

CONCLUSIONS

Our data support the notion that host factors are more important than isolate serotype in determining the severity and outcome of invasive pneumococcal disease and that these outcomes are unlikely to change in association with nonvaccine serotype infection in the post-conjugate vaccine era.

Nosocomial bloodstream infections caused by Streptococcus pneumoniae

A retrospective study of Streptococcus pneumoniae bacteraemia among adult patients in two large teaching hospitals in Spain identified 108 (10.6%) of 1,020 episodes as nosocomial pneumococcal bloodstream infections (NPBIs). Seventy-seven clinical records with sufficient data were available for analysis. The interval between admission and a positive blood culture was 3--135 days (median 17 days; interquartile range 8--27). The main underlying and predisposing conditions for NPBI were malignancy (31%), chronic obstructive pulmonary disease (28.6%), heart failure (16.9%), chronic renal failure (15.6%), liver cirrhosis (13%) and infection with human immunodeficiency virus (13%). Overall, 31.2% of patients developed severe sepsis, 11.7% septic shock, and 3.9% multi-organ failure. The main portals of entry were pneumonia (70.1%), meningitis (5.2%) and primary peritonitis (5.2%). Of the responsible serogroups, 78% were included in the 23-valent polysaccharide vaccine. Thirty-five (45.5%) patients died, with death considered to be related to the NPBI in 21 (27.3%) cases. Following multivariate analysis, factors that independently predicted death after adjusting for age were: ultimately fatal underlying disease (OR, 8.9; 95% CI, 0.8--94.3; p<0.001); rapidly fatal underlying disease (OR, 15.0; 95% CI, 2.8--81.3; p<0.001); heart failure (OR, 8.11; 95% CI, 1.1--60.8; p<0.03); inadequate empirical therapy (OR, 10.6; 95% CI, 1.2--97; p<0.003); a severe sepsis score (OR, 9.5; 95% CI, 1.9--47.0; p<0.001); and septic shock or multi-organ failure (OR, 63.7; 95% CI, 4.9--820.7; p<0.001). Adequate empirical therapy was an independent protective factor (OR, 0.05; 95% CI, 0.04--0.58; p<0.005), but the use of more than one antimicrobial agent was not.

Pneumococcal bacteremia in a single center in Argentina

OBJECTIVE

To determine the clinical and microbiologic characteristics of pneumococcal bacteremia at Sanatorio Mitre, Buenos Aires, Argentina.

METHODS

One-hundred-and-seven episodes of pneumococcal bacteremia were prospectively analyzed from 1993 to 1998. Demographics, clinical and microbiological variables were studied.

RESULTS

Eighty-one patients (76%) were adults and 26 children (24%). Most cases (98%) were acquired in the community. Seventy-nine patients (74%) had at least one underlying condition. The primary source of bacteremia was found in 91 patients (85%), the lungs being the most common source. Streptococcus pneumoniae was isolated from one sterile site other than the primary focus in 25 patients (23%). Eighty-five (79%) of the Streptococcus pneumoniae were susceptible to penicillin and 22 (21%) showed intermediate or high resistance to penicillin and 2% were additionally resistant to ceftriaxone. Initial antimicrobial therapy was appropriate in 95% of the cases. The overall mortality was 21%, however adults admitted to the intensive care unit (ICU) had higher mortality (81%). No patients under 14 years old died. Multivariate analysis showed that age and recovery of the organisms from a sterile site other than the primary focus were statistically significant predictors of mortality.

CONCLUSION

Bacteremic pneumococcal infections continue to be an important worldwide problem causing morbidity and high mortality despite supportive care and appropriate antimicrobial therapy.

Hospital-acquired pneumococcal bacteremia

To assess the most relevant features of hospital-acquired pneumococcal bacteremia, all cases of pneumococcal bacteremia at a single teaching hospital that occurred during 1988-2000 were prospectively studied. During this period, 374 cases of pneumococcal bacteremia were documented; 39 (10%) of these episodes were hospital acquired. Twenty-nine (74%) cases occurred during the period of December through May. Eleven (28%) of 39 patients had received antimicrobial agents in the month before the onset of bacteremia. All patients had underlying diseases that predisposed them to pneumococcal infection. The most common origin of infection was the respiratory tract, followed by the intra-abdominal region. Fifteen strains were fully susceptible to penicillin, and 20 were intermediately resistant. Only 25 strains were susceptible to erythromycin; all strains that were resistant to erythromycin were penicillin nonsusceptible. Eighteen (46%) of 39 patients died; the mortality rate related to infection was 39%.

Penicillin-resistant Streptococcus pneumoniae in metropolitan New York hospitals: case control study and molecular typing of resistant isolates

During the 4-month period from January to April, 1998, 476 patients with Streptococcus pneumoniae infections were detected in 12 metropolitan New York hospitals and 112 penicillin-resistant (PRP) isolates (24%) were identified in 11 institutions. A case control study of 100 patients with penicillin-resistant and susceptible pneumococci from four of the widely dispersed hospitals revealed a high incidence of underlying medical illnesses in adult patients (74%), a preponderance of patients with pneumonia (63%), and a majority of patients who had underlying risk factors for pneumonia or invasive disease (51%). In this limited case control study, no difference was noted between cases and controls regarding known risk factors for penicillin-resistant pneumococcal infections. The percentage of single-patient PRP isolates varied among individual hospitals but the mean percentages of PRP from the four participating University Medical Centers and seven community hospitals were similar: 26% and 22% respectively. By E-test, 60% and 26% were high-level penicillin and ceftriaxone resistant, respectively. Pulsed-field gel electrophoresis identified 26 chromosomal macrorestriction patterns among the 103 PRP isolates available for analysis, but almost half (50 isolates or 48%) of these belong to two drug-resistant internationally spread clones, SP(23)-1 and SP(9/14)-3, that were detected in all hospitals and were recovered from invasive and noninvasive sites in both children and adults.

In vitro susceptibility of recent clinical isolates of pneumococci to the investigational cephalosporin cefditoren

From February to June 2000, 2,597 isolates of Streptococcus pneumoniae were prospectively collected from 146 clinical laboratories across the United States (US) and tested to evaluate the in vitro activity of cefditoren, an investigational oral cephalosporin. In all, 2,492 isolates (96.0%) had a cefditoren MIC of 0.5 microg/mL or less, 74 isolates (2.8%) had an MIC of 1 microg/mL, 30 isolates (1.2%) had an MIC of 2 microg/mL, and 1 isolate (<0.1%) had an MIC of 4 microg/mL. Among the beta-lactams tested, the rank order of potency (MIC(90,) microg/mL) was cefditoren (0.5) > ceftriaxone (1) > amoxicillin-clavulanate (2) > cefuroxime (4) > cefprozil (8). Penicillin-resistant isolates (n = 443; 17.1%) were inhibited by lower concentrations (MIC(90,) microg/mL; MIC range,) of cefditoren (1; 0.03-4) than ceftriaxone (2; 0.25- > 2), amoxicillin-clavulanate (8; 0.5- > 8), cefuroxime (16; 2- > 16), and cefprozil (32; 2- > 32). Cefditoren MIC(90)s against cefuroxime-resistant (n = 640) and ceftriaxone-resistant (n = 89) isolates were 1 and 2 microg/mL, respectively. All isolates with reduced susceptibility to cefditoren (MIC, 2 or 4 microg/mL; n = 31) were resistant to penicillin, cefuroxime, and ceftriaxone. The potent in vitro activity of cefditoren against a recent US collection of pneumococci as demonstrated in this study supports its continued development for oral empiric therapy in outpatients with respiratory tract infections.

[Pneumococcal bacteremia in patients aged over 65 years. A study of 161 cases]

BACKGROUND

Here we present the results from a prospective analysis of pneumococcal bacteremia in patients older than 65 years, highlighting its differences with regard to patients aged from 15 to 64 years.

PATIENTS AND METHOD

We performed a prospective follow-up study (1988-1999) of patients older than 14 years whohad blood cultures positive for Streptococcus pneumoniae(S. pneumoniae). S. pneumoniae was identified using standard techniques. The antimicrobial sensitivity was analysed using the broth midrodilution technique and the serotype was studied in a reference laboratory. Statistical analyses were performed by means of the SPSS program, version 9.0.

RESULTS

161 out of 321 bacteremias occurred in 154 patients older than 65 years (86 males). Most frequent underlying diseases were chronic obstructive pulmonary disease (COPD), diabetes and neoplasia. Most patients had pneumonia, 29.7% with ilfiltrates involving at least two lung lobes. 9.4% of bacteremias were nosocomial. Among patients older than 65 years, 32.5% of strains were resistant to penicillin, compared to 21.7% in patients younger than 65 years(p = 0.034). Mortality in patients over 65 years was higher than in younger patients (25.3% vs 11.5%; p = 0.002) and mortality predictive factors in a multivariate analysis were age, nosocomial bacteremia and multi-lobe involvement (in patients with pneumonia).

CONCLUSIONS

There is a high incidence of pneumococcal bacteremia in patients over 65 years of age. These patients have a higher rate of resistance to penicillin and a higher mortality rate than younger patients. Mortality predictive factors are age, nosocomial bacteremia and multi-lobe involvement.