- Email: [email protected]
Comparison of clinical features, antimicrobial susceptibility, serotype distribution and outcomes of patients with hospital- and community-associated invasive pneumococcal disease
International Journal of Antimicrobial Agents 36 (2010) 119–123
Contents lists available at ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Comparison of clinical features, antimicrobial susceptibility, serotype distribution and outcomes of patients with hospital- and community-associated invasive pneumococcal disease Sheng-Hsiang Lin a , Wan-Hsiu Liao b , Chih-Cheng Lai c , Che-Kim Tan d , Chun-Hsing Liao e , Yu-Tsung Huang f , Cheng-Yi Wang c , Po-Ren Hsueh f,∗ a
Department of Internal Medicine, Taipei County Hospital, Taipei County, Taiwan Department of Family Medicine, Taipei County Hospital, Taipei County, Taiwan c Department of Internal Medicine, Cardinal Tien Hospital, Taipei County, Taiwan d Department of Intensive Care Medicine, Chi-Mei Medical Center, Tainan, Taiwan e Department of Internal Medicine, Far Eastern Memorial Hospital, Taipei County, Taiwan f Departments of Laboratory Medicine and Internal Medicine, National Taiwan University Hospital, National Taiwan University College of Medicine, Taipei, Taiwan b
a r t i c l e
i n f o
Article history: Received 2 February 2010 Accepted 16 April 2010 Keywords: Invasive pneumococcal disease Hospital-associated Community-associated Serotype Susceptibility Outcome
a b s t r a c t Hospital-associated invasive pneumococcal disease (HA-IPD) is infrequently reported. A retrospective surveillance of IPD in a medical centre in Taiwan was conducted from 2000 to 2008 to compare the clinical and microbiological characteristics of HA-IPD and community-associated IPD (CA-IPD). HA-IPD was identified in 37 patients, comprising 12.3% of the 302 hospitalised patients with IPD. Patients with HA-IPD were more likely to have solid-organ cancer (40.5% vs. 16.6%; P = 0.001) or to have received immunosuppressive therapy (56.8% vs. 26.8%; P < 0.001). The 30-day mortality rate of HA-IPD was significantly higher than that of CA-IPD (40.5% vs. 16.2%; P = 0.001). Age ≥65 years [odds ratio (OR) = 2.10; P = 0.033], HA-IPD (OR = 2.90; P = 0.009) and liver cirrhosis (OR = 3.19; P = 0.009) were independent predictors of 30-day mortality. No significant differences in serotype distribution or in susceptible rates to penicillin (18.2% vs. 32.6%; P = 0.14) and cefotaxime (60.6% vs. 67.8%; P = 0.53) were found between HA-IPD and CA-IPD isolates. Similar prevalences of the serotypes included in the pneumococcal vaccines were found in isolates from patients with HA-IPD and CA-IPD. Among patients with HA-IPD and CA-IPD, 26 (78.8%) and 172 (73.2%) (P = 0.45) had isolates of serotypes included in the 7-valent pneumococcal conjugate vaccine, and 30 (90.9%) and 224 (95.3%) (P = 0.96) had isolates of serotypes included in the 23-valent pneumococcal polysaccharide vaccine, respectively. In summary, this study found that HA-IPD and CA-IPD were not significantly different with regard to serotype distribution and antimicrobial susceptibility in Taiwan. Patients with HA-IPD have a higher mortality rate, and pneumococcal vaccination for patients at increased risk for HA-IPD should be encouraged. © 2010 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
1. Introduction Streptococcus pneumoniae is a well-known cause of communityacquired infections, and a substantial proportion of pneumococcal bacteraemias are hospital-acquired [1–6]. Nosocomial pneumococcal disease is associated with the emergence of antibiotic-resistant strains [7–10] and carries a higher mortality rate than communityassociated pneumococcal infection [1,4,11]. However, hospitalassociated pneumococcal disease has received little attention and thorough comparison of the characteristics of hospitalassociated and community-associated pneumococcal disease has been reported in only a few studies [1,10,12,13].
∗ Corresponding author. Tel.: +886 2 2312 3456x65355; fax: +886 2 2322 4263. E-mail address: [email protected] (P.-R. Hsueh).
Given the high prevalence of -lactam- and macrolide-resistant S. pneumoniae in Taiwan [14], studies to evaluate hospitalassociated infections caused by S. pneumoniae are well justified. This study compared the clinical characteristics, microbiological features and outcome of patients with hospital-associated invasive pneumococcal disease (HA-IPD) and community-associated invasive pneumococcal disease (CA-IPD).
2. Materials and methods 2.1. Setting and patients This retrospective cohort study included all hospitalised patients with IPD at National Taiwan University Hospital, a 2200bed tertiary care centre in northern Taiwan, during the 9-year
0924-8579/$ – see front matter © 2010 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/j.ijantimicag.2010.04.001
120
S.-H. Lin et al. / International Journal of Antimicrobial Agents 36 (2010) 119–123
period from January 2000 to December 2008. Charts were reviewed according to a pre-established protocol. 2.2. Definitions IPD was defined as the isolation of S. pneumoniae from a normally sterile site such as blood, cerebrospinal fluid (CSF), pleural fluid or ascites. Clinical diagnoses associated with IPD were classified as invasive pneumococcal pneumonia, meningitis, primary bacteraemia and other syndromes, as described in a previous report [15]. Pneumococcal meningitis was defined as the isolation of pneumococcus from CSF, or the clinical diagnosis of meningitis and isolation of pneumococcus from any sterile site. Invasive pneumococcal pneumonia was defined as the isolation of pneumococcus from pleural fluid, or a clinical diagnosis of pneumonia and isolation of pneumococcus from any sterile site. Primary pneumococcal bacteraemia was defined as isolation of pneumococcus from blood without an identified focus. IPD was defined as hospital-associated if the first positive sterile site culture was obtained >48 h after hospital admission or if the patient had been hospitalised for >2 days within 7 days of the first positive sterile site culture [1]. 2.3. Data collection The medical records of patients were reviewed to collect data regarding age, sex, date of admission, type of IPD diagnosis, Intensive Care Unit admission, septic shock, respiratory failure and death during the first 30 days after sterile site culture (all-cause mortality). The following co-morbid conditions were recorded: chronic lung disease (chronic obstructive pulmonary disease, asthma, bronchiectasis, pulmonary fibrosis, history of pulmonary tuberculosis); chronic heart disease (congenital heart disease, coronary heart disease, valvular heart disease, congestive heart failure); neurological disease (cerebrovascular disease, dementia, parkinsonism, epilepsy, cerebral palsy); diabetes; liver cirrhosis; chronic renal disease (chronic renal failure requiring dialysis, nephrotic syndrome); splenectomy or functional asplenia; human immunodeficiency virus (HIV) infection; solid-organ cancer; haematological malignancy; and immunosuppressive therapy (including chemotherapy, radiotherapy, transplantation and longterm use of systemic corticosteroids). 2.4. Microbiological studies Pneumococcal isolates were identified by recognition of typical colony morphology on trypticase soy agar supplemented with 5% sheep blood (BBL Microbiology Systems, Cockeysville, MD), Gram staining characteristics, susceptibility to ethylhydrocupreine hydrochloride (optochin) (Difco Laboratories, Detroit, MI) and bile solubility. Antimicrobial susceptibilities to erythromycin, clindamycin, tetracycline and chloramphenicol were determined by the disk diffusion method as described by the Clinical and Laboratory Standards Institute (CLSI) [16]. Minimum inhibitory concentrations (MICs) of penicillin, cefotaxime, levofloxacin and moxifloxacin were determined for available isolates using the agar dilution method and interpretation was categorised according to the 2009 CLSI guidelines [17]. Serotypes were determined for available isolates by latex agglutination (Pneumotest-Latex; Statens Serum Institut, Copenhagen, Denmark) according to the manufacturer’s instructions and were confirmed by Quellung reaction. 2.5. Statistical analysis Student’s t-test was used to compare continuous variables. Fisher’s exact test or 2 test was used to compare proportions.
Survival curves were calculated using Kaplan–Meier analysis and the log-rank test. Univariate and multivariate logistic regression analyses were used to analyse the association of clinical characteristics with 30-day mortality. Variables with a value of P < 0.1 were included in the multivariate analysis, and P-values of <0.05 in the multivariate analysis were considered to be statistically significant. All analyses were performed with SPSS version 10.0 (SPSS Inc., Chicago, IL). 3. Results 3.1. Patient characteristics During the 9-year study period, 37 patients with HA-IPD were identified, comprising 12.3% of all 302 hospitalised patients with IPD. Among patients with HA-IPD, the median time from admission to the index sterile site culture was 11 days (range 3–58 days). No clustering of IPD by ward or by contact with specific healthcare personnel was identified. Several characteristics of patients with HA-IPD differed from those patients with CA-IPD (Table 1). More patients with HA-IPD had an underlying medical condition (P = 0.047). This difference was due to a significantly greater proportions of patients with solid-organ cancer and immunosuppressive therapy. Significantly more patients with HA-IPD had primary bacteraemia (P < 0.001) and significantly more patients with CA-IPD had invasive pneumonia (P < 0.001). In the HA-IPD group, peritonitis was diagnosed in two patients (5.4%) and mycotic aortic aneurysm in one patient (2.7%). In the CA-IPD group, peritonitis was diagnosed in nine patients (3.4%), septic arthritis in five patients (1.9%), infective endocarditis in four patients (1.5%), acute otitis media in three patients (1.1%), biliary tract infection in two patients (0.8%), mycotic aortic aneurysm in one patient (0.4%) and pelvic abscess in one patient (0.4%).
Table 1 Comparison of characteristics of patients with hospital-associated (HA-) and community-associated (CA-) invasive pneumococcal disease (IPD)a . Characteristic Demographics Age (years) [mean (S.D.)] Age ≥65 years Age <5 years Male Underlying conditions Neurological disease Heart disease Chronic lung disease Liver cirrhosis Dialysis/nephrotic syndrome Diabetes mellitus Haematological malignancy Solid-organ cancer Immunosuppressive therapy Asplenia HIV infection ≥1 co-morbidity Clinical syndrome Meningitis Invasive pneumonia Other Primary bacteraemia Clinical course ICU stay Septic shock Respiratory failure
HA-IPD (n = 37)
CA-IPD (n = 265)
P-value
42.4 (28.3) 11 (29.7) 7 (18.9) 22 (59.5)
39.8 (30.9) 71 (26.8) 73 (27.5) 165 (62.3)
0.62 0.70 0.32 0.88
4 (10.8) 8 (21.6) 7 (18.9) 3 (8.1) 2 (5.4) 8 (21.6) 4 (10.8) 15 (40.5) 21 (56.8) 0 (0.0) 0 (0.0) 31 (83.8)
30 (11.3) 36 (13.6) 40 (15.1) 32 (12.1) 7 (2.6) 29 (10.9) 25 (9.4) 44 (16.6) 71 (26.8) 6 (2.3) 4 (1.5) 175 (66.0)
>0.99 0.29 0.72 0.67 0.30 0.11 0.77 0.001 <0.001 >0.99 >0.99 0.047
3 (8.1) 16 (43.2) 3 (8.1) 15 (40.5)
19 (7.2) 181 (68.3) 25 (9.4) 40 (15.1)
0.74 <0.001 >0.99 <0.001
19 (51.4) 17 (45.9) 17 (45.9)
124 (46.8) 86 (32.5) 88 (33.2)
0.73 0.15 0.18
S.D., standard deviation; HIV, human immunodeficiency virus; ICU, Intensive Care Unit. a Data are represented as n (%) of patients unless otherwise indicated.
S.-H. Lin et al. / International Journal of Antimicrobial Agents 36 (2010) 119–123
121
Fig. 1. Serotype distribution of pneumococcal isolates from patients with hospital-associated invasive pneumococcal disease (HA-IPD) and community-associated invasive pneumococcal disease (CA-IPD). * Indicated serotypes included in the 7-valent (PCV7), 10-valent (PCV10) and 13-valent (PCV13) pneumococcal conjugate vaccines and the 23-valent pneumococcal polysaccharide vaccine (PPV23). Other serotypes associated with CA-IPD included serotype 20 (0.9%), 22F (0.9%), 10A (0.4%), 1 (0.4%), 10B (0.4%), 17F (0.4%), 22A (0.4%) and 7F (0.4%). NT, non-typeable.
3.2. Serotype distribution Streptococcus pneumoniae isolates were available for serotyping in 33 HA-IPD patients and 235 CA-IPD patients. The serotype distribution of isolates from HA-IPD and CA-IPD patients is shown in Fig. 1. No significant difference in the prevalence of any of the serotypes was found between the two groups (all P > 0.05). Similar prevalences of the serotypes included in the pneumococcal
polysaccharide vaccines were found in isolates from patients with HA-IPD and CA-IPD. Among isolates from patients with HA-IPD and CA-IPD, 26 (78.8%) and 172 (73.2%) (P = 0.45) had serotypes included in the 7-valent pneumococcal conjugate vaccine (PCV7), 26 (78.8%) and 174 (74.0%) (P = 0.45) had isolates included in the 10-valent pneumococcal conjugate vaccine, 28 (84.8%) and 208 (88.5%) (P = 0.51) had isolates included in the 13-valent pneumococcal conjugate vaccine and 30 (90.9%) and 224 (95.3%) (P = 0.96)
Table 2 Comparison of the antimicrobial susceptibility of pneumococcal isolates from patients with hospital-associated (HA-) and community-associated (CA-) invasive pneumococcal disease (IPD). Antimicrobial agent By the agar dilution method Penicillin (n = 269) MIC50 (g/mL) MIC90 (g/mL) MIC range (g/mL) Susceptible rate (meningitis criteria) [n (%)] Susceptible rate (non-meningitis criteria) [n (%)]a Cefotaxime (n = 269) MIC50 (g/mL) MIC90 (g/mL) MIC range (g/mL) Susceptible rate (meningitis criteria) [n (%)]a Susceptible rate (non-meningitis criteria) [n (%)]a Levofloxacin (n = 269) MIC50 (g/mL) MIC90 (g/mL) MIC range (g/mL) Susceptible rate [n (%)]a Moxifloxacin (n = 269) MIC50 (g/mL) MIC90 (g/mL) MIC range (g/mL) Susceptible rate [n (%)]a By the disk diffusion method (susceptible rate) [n (%)]a Erythromycin (n = 302) Clindamycin (n = 302) Tetracycline (n = 302) Chloramphenicol (n = 298)
HA-IPD
CA-IPD
1.0 2.0 ≤0.03–4.0 6 (18.2) 32 (97.0)
0.5 2.0 ≤0.03–4.0 77 (32.6) 231 (97.9)
0.14 0.77
0.5 1.0 ≤0.03–4.0 20 (60.6) 30 (90.9)
0.5 1.0 ≤0.03–8.0 160 (67.8) 224 (94.9)
0.53 0.59
1.0 2.0 0.5–2.0 33 (100)
1.0 1.0 ≤0.25–8.0 235 (99.6)
>0.99
0.12 0.25 ≤0.03–0.25 33 (100)
0.12 0.25 ≤0.03–2.0 234 (99.2)
>0.99
4 (10.8) 16 (43.2) 9 (24.3) 25 (67.6)
37 (14.0) 92 (34.7) 44 (16.6) 203 (77.8)
0.80 0.36 0.25 0.24
MIC, minimum inhibitory concentration; MIC50/90 , MIC for 50% and 90% of the organisms, respectively. a According to the 2009 Clinical and Laboratory Standards Institute guidelines [17].
P-value
122
S.-H. Lin et al. / International Journal of Antimicrobial Agents 36 (2010) 119–123
neither acquisition of PNSSP (16.6% vs. 19.3%; P > 0.99) nor acquisition of cefotaxime-non-susceptible S. pneumoniae (20.0% vs. 19.2%; P > 0.99) were associated with a higher 30-day mortality rate. Factors significantly associated with death within 30 days after the first positive sterile site culture in the univariate analysis were age ≥65 years, HA-IPD, solid-organ cancer and immunosuppressive therapy, and in the multivariate analysis were age ≥65 years, HA-IPD and liver cirrhosis (Table 3).
4. Discussion
Fig. 2. Kaplan–Meier survival estimates of patients with community-associated invasive pneumococcal disease (CA-IPD) and hospital-associated invasive pneumococcal disease (HA-IPD) (hazard ratio = 2.81; 95% confidence interval 2.05–10.79; P < 0.001, log-rank test).
had isolates of serotypes included in the 23-valent pneumococcal polysaccharide vaccine (PPV23), respectively. 3.3. Antimicrobial susceptibility Antimicrobial susceptibilities of isolates from patients with HAIPD and CA-IPD are compared in Table 2. No significant difference in susceptible rates was found for any of the agents tested (all P > 0.05). Two cases of CA-IPD caused by isolates non-susceptible to fluoroquinolones (resistant to levofloxacin and intermediate to moxifloxacin in one patient, and intermediate to levofloxacin and susceptible to moxifloxacin in another) were found in 2007. The eight major serotypes arranged in order of decreasing penicillin susceptibility (percent susceptibility given in parentheses) by 2009 CLSI meningitis criteria were 3 (92.3%), 15B (55.6%), 19A (20%), 14 (18.2%), 6B (15.8%), 9V (12.5%), 19F (6.7%) and 23F (0%) and in order of decreasing cefotaxime susceptibility by 2009 CLSI meningitis criteria were 3 (100%), 15B (100%), 14 (76.6%), 6B (60.5%), 9V (50%), 19A (40%), 19F (36.7%) and 23F (29.7%). 3.4. Outcome The overall mortality rate at 30 days was significantly higher in patients with HA-IPD than in patients with CA-IPD (40.5% vs. 16.2%; P = 0.001). Kaplan–Meier survival plots for patients with HA-IPD and CA-IPD are shown in Fig. 2. By 2009 CLSI meningitis criteria for all isolates recovered from patients with IPD, neither acquisition of penicillin-non-susceptible S. pneumoniae (PNSSP) (20.4% vs. 16.9%; P = 0.61) nor acquisition of cefotaxime-non-susceptible S. pneumoniae (21.4% vs. 18.3%; P = 0.67) were associated with a higher 30-day mortality rate. By 2009 CLSI non-meningitis criteria for isolates recovered from patients with non-meningeal IPD,
This study comprehensively assessed the disease burden, risk factors, microbiological features and outcomes of HA-IPD in a medical centre in Taiwan. Some of the findings were similar to the results of previous studies. HA-IPD accounted for 12.3% of all IPD in this study, which is consistent with previous findings including those for episodes that developed 48–72 h after hospitalisation [1,2,6]. Previous studies found that more patients with nosocomial pneumococcal infection had underlying disease [10], and malignancy was the most important clinical risk factor for nosocomial pneumococcal bacteraemia [11,13]. In the present study, >80% of patients with HA-IPD had an underlying condition and >50% had solid-organ or haematological cancer, a significantly greater percentage than patients with CA-IPD. In addition, primary bacteraemia was a common manifestation of HA-IPD. These findings are in contrast to previous studies in which the majority of nosocomial pneumococcal infections were associated with pneumonia [6,10,11,13]. This may be attributable to differences in the epidemiological characteristics of the current patients, such as the high rate of immunosuppressive therapy in patients with HA-IPD. Previous studies found that nosocomial pneumococcal infection was associated with a higher mortality rate than communityassociated infection [1,3,4,11,13]. Consistent with these studies, this multivariate analysis showed that HA-IPD was associated with a significantly higher 30-day mortality rate than CA-IPD. Nosocomial IPD was not associated with increased antimicrobial resistance compared with CA-IPD in this study. Although two previous studies found that nosocomial acquisition was not a risk factor for the acquisition of PNSSP [3,8], several other studies found a significant difference in the distribution of PNSSP depending on whether the infection was nosocomial or community-associated [7,9,10]. The relatively small number of HA-IPD patients in this study may explain why this difference did not reach statistical significance. In addition, because Taiwan is an area of high prevalence of pneumococcal antibiotic resistance [14], the high selective antibiotic pressure that contributed to alterations in pneumococcal antibiotic sensitivity [7–10] of HA-IPD would likely be attenuated. Another possible reason for this discrepancy is the heterogeneity in the pathogenesis of HA-IPD. The infecting strain of S. pneumoniae in HA-IPD may be carried by the patient at the time of hospitalisation [18] but may also be spread from an asymptomatic carrier or from another patient with pneumococcal disease in the hospital [19].
Table 3 Logistic regression analysis of factors associated with 30-day mortality of invasive pneumococcal disease (IPD). Characteristic
Age ≥65 years Hospital-associated IPD Liver cirrhosis Heart disease Solid-organ cancer Immunosuppressive therapy ≥1 co-morbidity OR, odds ratio; CI, confidence interval.
Univariate
Multivariate
OR (95% CI)
P-value
OR (95% CI)
P-value
2.49 (1.37–4.52) 3.52 (1.69–7.33) 2.15 (0.98–4.68) 1.98 (0.96–4.09) 2.77 (1.46–5.25) 1.83 (1.01–3.31) 4.18 (1.82–9.61)
0.003 <0.001 0.055 0.063 0.002 0.046 <0.001
2.10 (1.06–4.17) 2.90 (1.31–6.42) 3.19 (1.34–7.58) 1.89 (0.80–4.47) 1.99 (0.89–4.43) 1.47 (0.68–3.22) Not adjusted
0.033 0.009 0.009 0.15 0.09 0.33
S.-H. Lin et al. / International Journal of Antimicrobial Agents 36 (2010) 119–123
The former case may resemble the antibiotic susceptibility profiles originating from the community, whilst the latter case may be more frequently associated with multidrug-resistant S. pneumoniae [20–22]. A number of studies have demonstrated geographic variation in the serotype distribution of pneumococcal isolates from nosocomial pneumococcal infection [1,2,5,6,10,13]. However, few studies have compared pneumococcal serotypes between patients with community-associated and nosocomial pneumococcal infection [10,13]. This study found that serotypes 23F and 19F were more frequent causes of HA-IPD than CA-IPD, although this difference was not significant. Similarly, Lee et al. [10] showed that isolates of serotypes 23F (33.3% vs. 11.9%) and 19F (31.3% vs. 23.9%) were more frequent causes of nosocomial pneumococcal infection. Serotypes 23F and 19F, which have the highest rates of PNSSP, were found to be the predominant strains in nosocomial pneumococcal transmission in Japan [23]. These serotypes were also responsible for the spread of pneumococcal strains with high -lactam resistance in Taiwan [14,24], including the international clones Taiwan23F , Spain23F and Taiwan19F . Further work is needed to establish the in-hospital transmission routes responsible for these infections so that more effective control policies can be developed. In Taiwan, PPV23 was introduced in 2001 and the cumulative coverage rate among people aged ≥75 years reached 41% in 2008; PCV7 was introduced in October 2005 and the cumulative coverage rate among children aged <5 years was 25.2% in 2008. Following the introduction of pneumococcal vaccines, a decline in the incidence of pneumococcal infections was documented [15,25]. Furthermore, our study showed that currently available pneumococcal vaccines would provide optimal efficacy both against HA-IPD and CA-IPD in Taiwan because >70% of the pneumococcal isolates from both groups were those included in the pneumococcal vaccines. When considered together with the high mortality, these data warrant increased efforts to improve the use of pneumococcal vaccines in persons who are at risk for HA-IPD [11]. This retrospective surveillance has several important limitations. Patients in this tertiary care centre in northern Taiwan are characterised by more complex disease syndromes as well as greater disease severity and incidence than a representative nationwide population from Taiwan. In addition, because genotyping or molecular typing of pneumococcal isolates was not performed, it was not possible to confirm the strain relatedness between isolates from patients with CA-IPD and HA-IPD. In conclusion, our results suggest that HA-IPD represents a distinct disease entity characterised by a higher rate of co-morbidity and fatality in Taiwan. The distributions of serotypes and antimicrobial resistances were not significantly different among patients with HA-IPD and CA-IPD. A high serotype coverage rate of pneumococcal vaccines was found both in HA-IPD and CA-IPD. These data suggest that patients who are at risk for HA-IPD would benefit more from pneumococcal vaccination. Funding: No funding sources. Competing interests: None declared. Ethical approval: Not required. References [1] Lyytikainen O, Klemets P, Ruutu P, Kaijalainen T, Rantala M, Ollgren J, et al. Defining the population-based burden of nosocomial pneumococcal bacteremia. Arch Intern Med 2007;167:1635–40.
123
[2] Canet JJ, Juan N, Xercavins M, Freixas N, Garau J. Hospital-acquired pneumococcal bacteremia. Clin Infect Dis 2002;35:697–702. [3] Maugein J, Guillemot D, Dupont MJ, Fosse T, Laurans G, Roussel-Delvallez M, et al. Clinical and microbiological epidemiology of Streptococcus pneumoniae bacteremia in eight French counties. Clin Microbiol Infect 2003;9: 280–8. [4] Farinas-Alvarez C, Farinas MC, Garcia-Palomo JD, Gonzalez-Ruiz M, FernandezMazarrasa C, Parra JA, et al. Prognostic factors for pneumococcal bacteremia in a university hospital. Eur J Clin Microbiol Infect Dis 2000;19:733–41. [5] Garcia-Leoni ME, Cercenado E, Rodeno P, Bernaldo de Quiros JC, MartinezHernandez D, Bouza E. Susceptibility of Streptococcus pneumoniae to penicillin: a prospective microbiological and clinical study. Clin Infect Dis 1992;14:427–35. [6] Bouza E, Pintado V, Rivera S, Blazquez R, Munoz P, Cercenado E, et al. Nosocomial bloodstream infections caused by Streptococcus pneumoniae. Clin Microbiol Infect 2005;11:919–24. [7] Nava JM, Bella F, Garau J, Lite J, Morera MA, Marti C, et al. Predictive factors for invasive disease due to penicillin-resistant Streptococcus pneumoniae: a population-based study. Clin Infect Dis 1994;19:884–90. [8] Vanderkooi OG, Low DE, Green K, Powis JE, McGeer A. Predicting antimicrobial resistance in invasive pneumococcal infections. Clin Infect Dis 2005;40:1288–97. [9] Bedos JP, Chevret S, Chastang C, Geslin P, Regnier B. Epidemiological features of and risk factors for infection by Streptococcus pneumoniae strains with diminished susceptibility to penicillin: findings of a French survey. Clin Infect Dis 1996;22:63–72. [10] Lee HJ, Park JY, Jang SH, Kim JH, Kim EC, Choi KW. High incidence of resistance to multiple antimicrobials in clinical isolates of Streptococcus pneumoniae from a university hospital in Korea. Clin Infect Dis 1995;20:826–35. [11] Rubins JB, Cheung S, Carson P, Rubins HB, Janoff EN. Identification of clinical risk factors for nosocomial pneumococcal bacteremia. Clin Infect Dis 1999;29:178–83. [12] Mylotte JM, Beam Jr TR. Comparison of community-acquired and nosocomial pneumococcal bacteremia. Am Rev Respir Dis 1981;123:265–8. [13] Alvarez S, Guarderas J, Shell CG, Holtsclaw-Berk S, Berk SL. Nosocomial pneumococcal bacteremia. Arch Intern Med 1986;146:1509–12. [14] Hsueh PR, Teng LJ, Wu TL, Yang D, Huang WK, Shyr JM, et al. Telithromycinand fluoroquinolone-resistant Streptococcus pneumoniae in Taiwan with high prevalence of resistance to macrolides and -lactams: SMART program 2001 data. Antimicrob Agents Chemother 2003;47:2145–51. [15] Liao WH, Lin SH, Lai CC, Tan CK, Liao CH, Huang YT, et al. Impact of pneumococcal vaccines on invasive pneumococcal disease in Taiwan. Eur J Clin Microbiol Infect Dis 2010;29:489–92. [16] National Committee for Clinical Laboratory Standards. Performance standard for antimicrobial disk susceptibility tests; approved standard. 5th ed. Document M2-A5. Wayne, PA: NCCLS; 1993. [17] Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing. Nineteenth informational supplement. Document M100-S19. Wayne, PA: CLSI; 2009. [18] Chen CJ, Huang YC, Su LH, Lin TY. Nasal carriage of Streptococcus pneumoniae in healthy children and adults in northern Taiwan. Diagn Microbiol Infect Dis 2007;59:265–9. [19] Nuorti JP, Butler JC, Crutcher JM, Guevara R, Welch D, Holder P, et al. An outbreak of multidrug-resistant pneumococcal pneumonia and bacteremia among unvaccinated nursing home residents. N Engl J Med 1998;338: 1861–8. [20] de Galan BE, van Tilburg PM, Sluijter M, Mol SJ, de Groot R, Hermans PW, et al. Hospital-related outbreak of infection with multidrug-resistant Streptococcus pneumoniae in the Netherlands. J Hosp Infect 1999;42:185–92. [21] Weiss K, Restieri C, Gauthier R, Laverdiere M, McGeer A, Davidson RJ, et al. A nosocomial outbreak of fluoroquinolone-resistant Streptococcus pneumoniae. Clin Infect Dis 2001;33:517–22. [22] Subramanian D, Sandoe JA, Keer V, Wilcox MH. Rapid spread of penicillinresistant Streptococcus pneumoniae among high-risk hospital inpatients and the role of molecular typing in outbreak confirmation. J Hosp Infect 2003;54:99–103. [23] Qin L, Masaki H, Watanabe K, Furumoto A, Watanabe H. Antimicrobial susceptibility and genetic characteristics of Streptococcus pneumoniae isolates indicating possible nosocomial transmission routes in a community hospital in Japan. J Clin Microbiol 2007;45:3701–6. [24] Hsieh YC, Chang KY, Huang YC, Lin HC, Ho YH, Huang LM, et al. Clonal spread of highly -lactam-resistant Streptococcus pneumoniae isolates in Taiwan. Antimicrob Agents Chemother 2008;52:2266–9. [25] Lin SH, Tan CK, Lai CC, Wang CY, Hsueh PR. Declining incidence of nonbacteremic pneumococcal pneumonia in hospitalized elderly patients at a tertiary care hospital after the introduction of pneumococcal vaccines in Taiwan, 2004–2008. J Am Geriatr Soc 2010;58:195–6.
Contents lists available at ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Comparison of clinical features, antimicrobial susceptibility, serotype distribution and outcomes of patients with hospital- and community-associated invasive pneumococcal disease Sheng-Hsiang Lin a , Wan-Hsiu Liao b , Chih-Cheng Lai c , Che-Kim Tan d , Chun-Hsing Liao e , Yu-Tsung Huang f , Cheng-Yi Wang c , Po-Ren Hsueh f,∗ a
Department of Internal Medicine, Taipei County Hospital, Taipei County, Taiwan Department of Family Medicine, Taipei County Hospital, Taipei County, Taiwan c Department of Internal Medicine, Cardinal Tien Hospital, Taipei County, Taiwan d Department of Intensive Care Medicine, Chi-Mei Medical Center, Tainan, Taiwan e Department of Internal Medicine, Far Eastern Memorial Hospital, Taipei County, Taiwan f Departments of Laboratory Medicine and Internal Medicine, National Taiwan University Hospital, National Taiwan University College of Medicine, Taipei, Taiwan b
a r t i c l e
i n f o
Article history: Received 2 February 2010 Accepted 16 April 2010 Keywords: Invasive pneumococcal disease Hospital-associated Community-associated Serotype Susceptibility Outcome
a b s t r a c t Hospital-associated invasive pneumococcal disease (HA-IPD) is infrequently reported. A retrospective surveillance of IPD in a medical centre in Taiwan was conducted from 2000 to 2008 to compare the clinical and microbiological characteristics of HA-IPD and community-associated IPD (CA-IPD). HA-IPD was identified in 37 patients, comprising 12.3% of the 302 hospitalised patients with IPD. Patients with HA-IPD were more likely to have solid-organ cancer (40.5% vs. 16.6%; P = 0.001) or to have received immunosuppressive therapy (56.8% vs. 26.8%; P < 0.001). The 30-day mortality rate of HA-IPD was significantly higher than that of CA-IPD (40.5% vs. 16.2%; P = 0.001). Age ≥65 years [odds ratio (OR) = 2.10; P = 0.033], HA-IPD (OR = 2.90; P = 0.009) and liver cirrhosis (OR = 3.19; P = 0.009) were independent predictors of 30-day mortality. No significant differences in serotype distribution or in susceptible rates to penicillin (18.2% vs. 32.6%; P = 0.14) and cefotaxime (60.6% vs. 67.8%; P = 0.53) were found between HA-IPD and CA-IPD isolates. Similar prevalences of the serotypes included in the pneumococcal vaccines were found in isolates from patients with HA-IPD and CA-IPD. Among patients with HA-IPD and CA-IPD, 26 (78.8%) and 172 (73.2%) (P = 0.45) had isolates of serotypes included in the 7-valent pneumococcal conjugate vaccine, and 30 (90.9%) and 224 (95.3%) (P = 0.96) had isolates of serotypes included in the 23-valent pneumococcal polysaccharide vaccine, respectively. In summary, this study found that HA-IPD and CA-IPD were not significantly different with regard to serotype distribution and antimicrobial susceptibility in Taiwan. Patients with HA-IPD have a higher mortality rate, and pneumococcal vaccination for patients at increased risk for HA-IPD should be encouraged. © 2010 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
1. Introduction Streptococcus pneumoniae is a well-known cause of communityacquired infections, and a substantial proportion of pneumococcal bacteraemias are hospital-acquired [1–6]. Nosocomial pneumococcal disease is associated with the emergence of antibiotic-resistant strains [7–10] and carries a higher mortality rate than communityassociated pneumococcal infection [1,4,11]. However, hospitalassociated pneumococcal disease has received little attention and thorough comparison of the characteristics of hospitalassociated and community-associated pneumococcal disease has been reported in only a few studies [1,10,12,13].
∗ Corresponding author. Tel.: +886 2 2312 3456x65355; fax: +886 2 2322 4263. E-mail address: [email protected] (P.-R. Hsueh).
Given the high prevalence of -lactam- and macrolide-resistant S. pneumoniae in Taiwan [14], studies to evaluate hospitalassociated infections caused by S. pneumoniae are well justified. This study compared the clinical characteristics, microbiological features and outcome of patients with hospital-associated invasive pneumococcal disease (HA-IPD) and community-associated invasive pneumococcal disease (CA-IPD).
2. Materials and methods 2.1. Setting and patients This retrospective cohort study included all hospitalised patients with IPD at National Taiwan University Hospital, a 2200bed tertiary care centre in northern Taiwan, during the 9-year
0924-8579/$ – see front matter © 2010 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/j.ijantimicag.2010.04.001
120
S.-H. Lin et al. / International Journal of Antimicrobial Agents 36 (2010) 119–123
period from January 2000 to December 2008. Charts were reviewed according to a pre-established protocol. 2.2. Definitions IPD was defined as the isolation of S. pneumoniae from a normally sterile site such as blood, cerebrospinal fluid (CSF), pleural fluid or ascites. Clinical diagnoses associated with IPD were classified as invasive pneumococcal pneumonia, meningitis, primary bacteraemia and other syndromes, as described in a previous report [15]. Pneumococcal meningitis was defined as the isolation of pneumococcus from CSF, or the clinical diagnosis of meningitis and isolation of pneumococcus from any sterile site. Invasive pneumococcal pneumonia was defined as the isolation of pneumococcus from pleural fluid, or a clinical diagnosis of pneumonia and isolation of pneumococcus from any sterile site. Primary pneumococcal bacteraemia was defined as isolation of pneumococcus from blood without an identified focus. IPD was defined as hospital-associated if the first positive sterile site culture was obtained >48 h after hospital admission or if the patient had been hospitalised for >2 days within 7 days of the first positive sterile site culture [1]. 2.3. Data collection The medical records of patients were reviewed to collect data regarding age, sex, date of admission, type of IPD diagnosis, Intensive Care Unit admission, septic shock, respiratory failure and death during the first 30 days after sterile site culture (all-cause mortality). The following co-morbid conditions were recorded: chronic lung disease (chronic obstructive pulmonary disease, asthma, bronchiectasis, pulmonary fibrosis, history of pulmonary tuberculosis); chronic heart disease (congenital heart disease, coronary heart disease, valvular heart disease, congestive heart failure); neurological disease (cerebrovascular disease, dementia, parkinsonism, epilepsy, cerebral palsy); diabetes; liver cirrhosis; chronic renal disease (chronic renal failure requiring dialysis, nephrotic syndrome); splenectomy or functional asplenia; human immunodeficiency virus (HIV) infection; solid-organ cancer; haematological malignancy; and immunosuppressive therapy (including chemotherapy, radiotherapy, transplantation and longterm use of systemic corticosteroids). 2.4. Microbiological studies Pneumococcal isolates were identified by recognition of typical colony morphology on trypticase soy agar supplemented with 5% sheep blood (BBL Microbiology Systems, Cockeysville, MD), Gram staining characteristics, susceptibility to ethylhydrocupreine hydrochloride (optochin) (Difco Laboratories, Detroit, MI) and bile solubility. Antimicrobial susceptibilities to erythromycin, clindamycin, tetracycline and chloramphenicol were determined by the disk diffusion method as described by the Clinical and Laboratory Standards Institute (CLSI) [16]. Minimum inhibitory concentrations (MICs) of penicillin, cefotaxime, levofloxacin and moxifloxacin were determined for available isolates using the agar dilution method and interpretation was categorised according to the 2009 CLSI guidelines [17]. Serotypes were determined for available isolates by latex agglutination (Pneumotest-Latex; Statens Serum Institut, Copenhagen, Denmark) according to the manufacturer’s instructions and were confirmed by Quellung reaction. 2.5. Statistical analysis Student’s t-test was used to compare continuous variables. Fisher’s exact test or 2 test was used to compare proportions.
Survival curves were calculated using Kaplan–Meier analysis and the log-rank test. Univariate and multivariate logistic regression analyses were used to analyse the association of clinical characteristics with 30-day mortality. Variables with a value of P < 0.1 were included in the multivariate analysis, and P-values of <0.05 in the multivariate analysis were considered to be statistically significant. All analyses were performed with SPSS version 10.0 (SPSS Inc., Chicago, IL). 3. Results 3.1. Patient characteristics During the 9-year study period, 37 patients with HA-IPD were identified, comprising 12.3% of all 302 hospitalised patients with IPD. Among patients with HA-IPD, the median time from admission to the index sterile site culture was 11 days (range 3–58 days). No clustering of IPD by ward or by contact with specific healthcare personnel was identified. Several characteristics of patients with HA-IPD differed from those patients with CA-IPD (Table 1). More patients with HA-IPD had an underlying medical condition (P = 0.047). This difference was due to a significantly greater proportions of patients with solid-organ cancer and immunosuppressive therapy. Significantly more patients with HA-IPD had primary bacteraemia (P < 0.001) and significantly more patients with CA-IPD had invasive pneumonia (P < 0.001). In the HA-IPD group, peritonitis was diagnosed in two patients (5.4%) and mycotic aortic aneurysm in one patient (2.7%). In the CA-IPD group, peritonitis was diagnosed in nine patients (3.4%), septic arthritis in five patients (1.9%), infective endocarditis in four patients (1.5%), acute otitis media in three patients (1.1%), biliary tract infection in two patients (0.8%), mycotic aortic aneurysm in one patient (0.4%) and pelvic abscess in one patient (0.4%).
Table 1 Comparison of characteristics of patients with hospital-associated (HA-) and community-associated (CA-) invasive pneumococcal disease (IPD)a . Characteristic Demographics Age (years) [mean (S.D.)] Age ≥65 years Age <5 years Male Underlying conditions Neurological disease Heart disease Chronic lung disease Liver cirrhosis Dialysis/nephrotic syndrome Diabetes mellitus Haematological malignancy Solid-organ cancer Immunosuppressive therapy Asplenia HIV infection ≥1 co-morbidity Clinical syndrome Meningitis Invasive pneumonia Other Primary bacteraemia Clinical course ICU stay Septic shock Respiratory failure
HA-IPD (n = 37)
CA-IPD (n = 265)
P-value
42.4 (28.3) 11 (29.7) 7 (18.9) 22 (59.5)
39.8 (30.9) 71 (26.8) 73 (27.5) 165 (62.3)
0.62 0.70 0.32 0.88
4 (10.8) 8 (21.6) 7 (18.9) 3 (8.1) 2 (5.4) 8 (21.6) 4 (10.8) 15 (40.5) 21 (56.8) 0 (0.0) 0 (0.0) 31 (83.8)
30 (11.3) 36 (13.6) 40 (15.1) 32 (12.1) 7 (2.6) 29 (10.9) 25 (9.4) 44 (16.6) 71 (26.8) 6 (2.3) 4 (1.5) 175 (66.0)
>0.99 0.29 0.72 0.67 0.30 0.11 0.77 0.001 <0.001 >0.99 >0.99 0.047
3 (8.1) 16 (43.2) 3 (8.1) 15 (40.5)
19 (7.2) 181 (68.3) 25 (9.4) 40 (15.1)
0.74 <0.001 >0.99 <0.001
19 (51.4) 17 (45.9) 17 (45.9)
124 (46.8) 86 (32.5) 88 (33.2)
0.73 0.15 0.18
S.D., standard deviation; HIV, human immunodeficiency virus; ICU, Intensive Care Unit. a Data are represented as n (%) of patients unless otherwise indicated.
S.-H. Lin et al. / International Journal of Antimicrobial Agents 36 (2010) 119–123
121
Fig. 1. Serotype distribution of pneumococcal isolates from patients with hospital-associated invasive pneumococcal disease (HA-IPD) and community-associated invasive pneumococcal disease (CA-IPD). * Indicated serotypes included in the 7-valent (PCV7), 10-valent (PCV10) and 13-valent (PCV13) pneumococcal conjugate vaccines and the 23-valent pneumococcal polysaccharide vaccine (PPV23). Other serotypes associated with CA-IPD included serotype 20 (0.9%), 22F (0.9%), 10A (0.4%), 1 (0.4%), 10B (0.4%), 17F (0.4%), 22A (0.4%) and 7F (0.4%). NT, non-typeable.
3.2. Serotype distribution Streptococcus pneumoniae isolates were available for serotyping in 33 HA-IPD patients and 235 CA-IPD patients. The serotype distribution of isolates from HA-IPD and CA-IPD patients is shown in Fig. 1. No significant difference in the prevalence of any of the serotypes was found between the two groups (all P > 0.05). Similar prevalences of the serotypes included in the pneumococcal
polysaccharide vaccines were found in isolates from patients with HA-IPD and CA-IPD. Among isolates from patients with HA-IPD and CA-IPD, 26 (78.8%) and 172 (73.2%) (P = 0.45) had serotypes included in the 7-valent pneumococcal conjugate vaccine (PCV7), 26 (78.8%) and 174 (74.0%) (P = 0.45) had isolates included in the 10-valent pneumococcal conjugate vaccine, 28 (84.8%) and 208 (88.5%) (P = 0.51) had isolates included in the 13-valent pneumococcal conjugate vaccine and 30 (90.9%) and 224 (95.3%) (P = 0.96)
Table 2 Comparison of the antimicrobial susceptibility of pneumococcal isolates from patients with hospital-associated (HA-) and community-associated (CA-) invasive pneumococcal disease (IPD). Antimicrobial agent By the agar dilution method Penicillin (n = 269) MIC50 (g/mL) MIC90 (g/mL) MIC range (g/mL) Susceptible rate (meningitis criteria) [n (%)] Susceptible rate (non-meningitis criteria) [n (%)]a Cefotaxime (n = 269) MIC50 (g/mL) MIC90 (g/mL) MIC range (g/mL) Susceptible rate (meningitis criteria) [n (%)]a Susceptible rate (non-meningitis criteria) [n (%)]a Levofloxacin (n = 269) MIC50 (g/mL) MIC90 (g/mL) MIC range (g/mL) Susceptible rate [n (%)]a Moxifloxacin (n = 269) MIC50 (g/mL) MIC90 (g/mL) MIC range (g/mL) Susceptible rate [n (%)]a By the disk diffusion method (susceptible rate) [n (%)]a Erythromycin (n = 302) Clindamycin (n = 302) Tetracycline (n = 302) Chloramphenicol (n = 298)
HA-IPD
CA-IPD
1.0 2.0 ≤0.03–4.0 6 (18.2) 32 (97.0)
0.5 2.0 ≤0.03–4.0 77 (32.6) 231 (97.9)
0.14 0.77
0.5 1.0 ≤0.03–4.0 20 (60.6) 30 (90.9)
0.5 1.0 ≤0.03–8.0 160 (67.8) 224 (94.9)
0.53 0.59
1.0 2.0 0.5–2.0 33 (100)
1.0 1.0 ≤0.25–8.0 235 (99.6)
>0.99
0.12 0.25 ≤0.03–0.25 33 (100)
0.12 0.25 ≤0.03–2.0 234 (99.2)
>0.99
4 (10.8) 16 (43.2) 9 (24.3) 25 (67.6)
37 (14.0) 92 (34.7) 44 (16.6) 203 (77.8)
0.80 0.36 0.25 0.24
MIC, minimum inhibitory concentration; MIC50/90 , MIC for 50% and 90% of the organisms, respectively. a According to the 2009 Clinical and Laboratory Standards Institute guidelines [17].
P-value
122
S.-H. Lin et al. / International Journal of Antimicrobial Agents 36 (2010) 119–123
neither acquisition of PNSSP (16.6% vs. 19.3%; P > 0.99) nor acquisition of cefotaxime-non-susceptible S. pneumoniae (20.0% vs. 19.2%; P > 0.99) were associated with a higher 30-day mortality rate. Factors significantly associated with death within 30 days after the first positive sterile site culture in the univariate analysis were age ≥65 years, HA-IPD, solid-organ cancer and immunosuppressive therapy, and in the multivariate analysis were age ≥65 years, HA-IPD and liver cirrhosis (Table 3).
4. Discussion
Fig. 2. Kaplan–Meier survival estimates of patients with community-associated invasive pneumococcal disease (CA-IPD) and hospital-associated invasive pneumococcal disease (HA-IPD) (hazard ratio = 2.81; 95% confidence interval 2.05–10.79; P < 0.001, log-rank test).
had isolates of serotypes included in the 23-valent pneumococcal polysaccharide vaccine (PPV23), respectively. 3.3. Antimicrobial susceptibility Antimicrobial susceptibilities of isolates from patients with HAIPD and CA-IPD are compared in Table 2. No significant difference in susceptible rates was found for any of the agents tested (all P > 0.05). Two cases of CA-IPD caused by isolates non-susceptible to fluoroquinolones (resistant to levofloxacin and intermediate to moxifloxacin in one patient, and intermediate to levofloxacin and susceptible to moxifloxacin in another) were found in 2007. The eight major serotypes arranged in order of decreasing penicillin susceptibility (percent susceptibility given in parentheses) by 2009 CLSI meningitis criteria were 3 (92.3%), 15B (55.6%), 19A (20%), 14 (18.2%), 6B (15.8%), 9V (12.5%), 19F (6.7%) and 23F (0%) and in order of decreasing cefotaxime susceptibility by 2009 CLSI meningitis criteria were 3 (100%), 15B (100%), 14 (76.6%), 6B (60.5%), 9V (50%), 19A (40%), 19F (36.7%) and 23F (29.7%). 3.4. Outcome The overall mortality rate at 30 days was significantly higher in patients with HA-IPD than in patients with CA-IPD (40.5% vs. 16.2%; P = 0.001). Kaplan–Meier survival plots for patients with HA-IPD and CA-IPD are shown in Fig. 2. By 2009 CLSI meningitis criteria for all isolates recovered from patients with IPD, neither acquisition of penicillin-non-susceptible S. pneumoniae (PNSSP) (20.4% vs. 16.9%; P = 0.61) nor acquisition of cefotaxime-non-susceptible S. pneumoniae (21.4% vs. 18.3%; P = 0.67) were associated with a higher 30-day mortality rate. By 2009 CLSI non-meningitis criteria for isolates recovered from patients with non-meningeal IPD,
This study comprehensively assessed the disease burden, risk factors, microbiological features and outcomes of HA-IPD in a medical centre in Taiwan. Some of the findings were similar to the results of previous studies. HA-IPD accounted for 12.3% of all IPD in this study, which is consistent with previous findings including those for episodes that developed 48–72 h after hospitalisation [1,2,6]. Previous studies found that more patients with nosocomial pneumococcal infection had underlying disease [10], and malignancy was the most important clinical risk factor for nosocomial pneumococcal bacteraemia [11,13]. In the present study, >80% of patients with HA-IPD had an underlying condition and >50% had solid-organ or haematological cancer, a significantly greater percentage than patients with CA-IPD. In addition, primary bacteraemia was a common manifestation of HA-IPD. These findings are in contrast to previous studies in which the majority of nosocomial pneumococcal infections were associated with pneumonia [6,10,11,13]. This may be attributable to differences in the epidemiological characteristics of the current patients, such as the high rate of immunosuppressive therapy in patients with HA-IPD. Previous studies found that nosocomial pneumococcal infection was associated with a higher mortality rate than communityassociated infection [1,3,4,11,13]. Consistent with these studies, this multivariate analysis showed that HA-IPD was associated with a significantly higher 30-day mortality rate than CA-IPD. Nosocomial IPD was not associated with increased antimicrobial resistance compared with CA-IPD in this study. Although two previous studies found that nosocomial acquisition was not a risk factor for the acquisition of PNSSP [3,8], several other studies found a significant difference in the distribution of PNSSP depending on whether the infection was nosocomial or community-associated [7,9,10]. The relatively small number of HA-IPD patients in this study may explain why this difference did not reach statistical significance. In addition, because Taiwan is an area of high prevalence of pneumococcal antibiotic resistance [14], the high selective antibiotic pressure that contributed to alterations in pneumococcal antibiotic sensitivity [7–10] of HA-IPD would likely be attenuated. Another possible reason for this discrepancy is the heterogeneity in the pathogenesis of HA-IPD. The infecting strain of S. pneumoniae in HA-IPD may be carried by the patient at the time of hospitalisation [18] but may also be spread from an asymptomatic carrier or from another patient with pneumococcal disease in the hospital [19].
Table 3 Logistic regression analysis of factors associated with 30-day mortality of invasive pneumococcal disease (IPD). Characteristic
Age ≥65 years Hospital-associated IPD Liver cirrhosis Heart disease Solid-organ cancer Immunosuppressive therapy ≥1 co-morbidity OR, odds ratio; CI, confidence interval.
Univariate
Multivariate
OR (95% CI)
P-value
OR (95% CI)
P-value
2.49 (1.37–4.52) 3.52 (1.69–7.33) 2.15 (0.98–4.68) 1.98 (0.96–4.09) 2.77 (1.46–5.25) 1.83 (1.01–3.31) 4.18 (1.82–9.61)
0.003 <0.001 0.055 0.063 0.002 0.046 <0.001
2.10 (1.06–4.17) 2.90 (1.31–6.42) 3.19 (1.34–7.58) 1.89 (0.80–4.47) 1.99 (0.89–4.43) 1.47 (0.68–3.22) Not adjusted
0.033 0.009 0.009 0.15 0.09 0.33
S.-H. Lin et al. / International Journal of Antimicrobial Agents 36 (2010) 119–123
The former case may resemble the antibiotic susceptibility profiles originating from the community, whilst the latter case may be more frequently associated with multidrug-resistant S. pneumoniae [20–22]. A number of studies have demonstrated geographic variation in the serotype distribution of pneumococcal isolates from nosocomial pneumococcal infection [1,2,5,6,10,13]. However, few studies have compared pneumococcal serotypes between patients with community-associated and nosocomial pneumococcal infection [10,13]. This study found that serotypes 23F and 19F were more frequent causes of HA-IPD than CA-IPD, although this difference was not significant. Similarly, Lee et al. [10] showed that isolates of serotypes 23F (33.3% vs. 11.9%) and 19F (31.3% vs. 23.9%) were more frequent causes of nosocomial pneumococcal infection. Serotypes 23F and 19F, which have the highest rates of PNSSP, were found to be the predominant strains in nosocomial pneumococcal transmission in Japan [23]. These serotypes were also responsible for the spread of pneumococcal strains with high -lactam resistance in Taiwan [14,24], including the international clones Taiwan23F , Spain23F and Taiwan19F . Further work is needed to establish the in-hospital transmission routes responsible for these infections so that more effective control policies can be developed. In Taiwan, PPV23 was introduced in 2001 and the cumulative coverage rate among people aged ≥75 years reached 41% in 2008; PCV7 was introduced in October 2005 and the cumulative coverage rate among children aged <5 years was 25.2% in 2008. Following the introduction of pneumococcal vaccines, a decline in the incidence of pneumococcal infections was documented [15,25]. Furthermore, our study showed that currently available pneumococcal vaccines would provide optimal efficacy both against HA-IPD and CA-IPD in Taiwan because >70% of the pneumococcal isolates from both groups were those included in the pneumococcal vaccines. When considered together with the high mortality, these data warrant increased efforts to improve the use of pneumococcal vaccines in persons who are at risk for HA-IPD [11]. This retrospective surveillance has several important limitations. Patients in this tertiary care centre in northern Taiwan are characterised by more complex disease syndromes as well as greater disease severity and incidence than a representative nationwide population from Taiwan. In addition, because genotyping or molecular typing of pneumococcal isolates was not performed, it was not possible to confirm the strain relatedness between isolates from patients with CA-IPD and HA-IPD. In conclusion, our results suggest that HA-IPD represents a distinct disease entity characterised by a higher rate of co-morbidity and fatality in Taiwan. The distributions of serotypes and antimicrobial resistances were not significantly different among patients with HA-IPD and CA-IPD. A high serotype coverage rate of pneumococcal vaccines was found both in HA-IPD and CA-IPD. These data suggest that patients who are at risk for HA-IPD would benefit more from pneumococcal vaccination. Funding: No funding sources. Competing interests: None declared. Ethical approval: Not required. References [1] Lyytikainen O, Klemets P, Ruutu P, Kaijalainen T, Rantala M, Ollgren J, et al. Defining the population-based burden of nosocomial pneumococcal bacteremia. Arch Intern Med 2007;167:1635–40.
123
[2] Canet JJ, Juan N, Xercavins M, Freixas N, Garau J. Hospital-acquired pneumococcal bacteremia. Clin Infect Dis 2002;35:697–702. [3] Maugein J, Guillemot D, Dupont MJ, Fosse T, Laurans G, Roussel-Delvallez M, et al. Clinical and microbiological epidemiology of Streptococcus pneumoniae bacteremia in eight French counties. Clin Microbiol Infect 2003;9: 280–8. [4] Farinas-Alvarez C, Farinas MC, Garcia-Palomo JD, Gonzalez-Ruiz M, FernandezMazarrasa C, Parra JA, et al. Prognostic factors for pneumococcal bacteremia in a university hospital. Eur J Clin Microbiol Infect Dis 2000;19:733–41. [5] Garcia-Leoni ME, Cercenado E, Rodeno P, Bernaldo de Quiros JC, MartinezHernandez D, Bouza E. Susceptibility of Streptococcus pneumoniae to penicillin: a prospective microbiological and clinical study. Clin Infect Dis 1992;14:427–35. [6] Bouza E, Pintado V, Rivera S, Blazquez R, Munoz P, Cercenado E, et al. Nosocomial bloodstream infections caused by Streptococcus pneumoniae. Clin Microbiol Infect 2005;11:919–24. [7] Nava JM, Bella F, Garau J, Lite J, Morera MA, Marti C, et al. Predictive factors for invasive disease due to penicillin-resistant Streptococcus pneumoniae: a population-based study. Clin Infect Dis 1994;19:884–90. [8] Vanderkooi OG, Low DE, Green K, Powis JE, McGeer A. Predicting antimicrobial resistance in invasive pneumococcal infections. Clin Infect Dis 2005;40:1288–97. [9] Bedos JP, Chevret S, Chastang C, Geslin P, Regnier B. Epidemiological features of and risk factors for infection by Streptococcus pneumoniae strains with diminished susceptibility to penicillin: findings of a French survey. Clin Infect Dis 1996;22:63–72. [10] Lee HJ, Park JY, Jang SH, Kim JH, Kim EC, Choi KW. High incidence of resistance to multiple antimicrobials in clinical isolates of Streptococcus pneumoniae from a university hospital in Korea. Clin Infect Dis 1995;20:826–35. [11] Rubins JB, Cheung S, Carson P, Rubins HB, Janoff EN. Identification of clinical risk factors for nosocomial pneumococcal bacteremia. Clin Infect Dis 1999;29:178–83. [12] Mylotte JM, Beam Jr TR. Comparison of community-acquired and nosocomial pneumococcal bacteremia. Am Rev Respir Dis 1981;123:265–8. [13] Alvarez S, Guarderas J, Shell CG, Holtsclaw-Berk S, Berk SL. Nosocomial pneumococcal bacteremia. Arch Intern Med 1986;146:1509–12. [14] Hsueh PR, Teng LJ, Wu TL, Yang D, Huang WK, Shyr JM, et al. Telithromycinand fluoroquinolone-resistant Streptococcus pneumoniae in Taiwan with high prevalence of resistance to macrolides and -lactams: SMART program 2001 data. Antimicrob Agents Chemother 2003;47:2145–51. [15] Liao WH, Lin SH, Lai CC, Tan CK, Liao CH, Huang YT, et al. Impact of pneumococcal vaccines on invasive pneumococcal disease in Taiwan. Eur J Clin Microbiol Infect Dis 2010;29:489–92. [16] National Committee for Clinical Laboratory Standards. Performance standard for antimicrobial disk susceptibility tests; approved standard. 5th ed. Document M2-A5. Wayne, PA: NCCLS; 1993. [17] Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing. Nineteenth informational supplement. Document M100-S19. Wayne, PA: CLSI; 2009. [18] Chen CJ, Huang YC, Su LH, Lin TY. Nasal carriage of Streptococcus pneumoniae in healthy children and adults in northern Taiwan. Diagn Microbiol Infect Dis 2007;59:265–9. [19] Nuorti JP, Butler JC, Crutcher JM, Guevara R, Welch D, Holder P, et al. An outbreak of multidrug-resistant pneumococcal pneumonia and bacteremia among unvaccinated nursing home residents. N Engl J Med 1998;338: 1861–8. [20] de Galan BE, van Tilburg PM, Sluijter M, Mol SJ, de Groot R, Hermans PW, et al. Hospital-related outbreak of infection with multidrug-resistant Streptococcus pneumoniae in the Netherlands. J Hosp Infect 1999;42:185–92. [21] Weiss K, Restieri C, Gauthier R, Laverdiere M, McGeer A, Davidson RJ, et al. A nosocomial outbreak of fluoroquinolone-resistant Streptococcus pneumoniae. Clin Infect Dis 2001;33:517–22. [22] Subramanian D, Sandoe JA, Keer V, Wilcox MH. Rapid spread of penicillinresistant Streptococcus pneumoniae among high-risk hospital inpatients and the role of molecular typing in outbreak confirmation. J Hosp Infect 2003;54:99–103. [23] Qin L, Masaki H, Watanabe K, Furumoto A, Watanabe H. Antimicrobial susceptibility and genetic characteristics of Streptococcus pneumoniae isolates indicating possible nosocomial transmission routes in a community hospital in Japan. J Clin Microbiol 2007;45:3701–6. [24] Hsieh YC, Chang KY, Huang YC, Lin HC, Ho YH, Huang LM, et al. Clonal spread of highly -lactam-resistant Streptococcus pneumoniae isolates in Taiwan. Antimicrob Agents Chemother 2008;52:2266–9. [25] Lin SH, Tan CK, Lai CC, Wang CY, Hsueh PR. Declining incidence of nonbacteremic pneumococcal pneumonia in hospitalized elderly patients at a tertiary care hospital after the introduction of pneumococcal vaccines in Taiwan, 2004–2008. J Am Geriatr Soc 2010;58:195–6.