American Journal of Respiratory and Critical Care Medicine

The aim of the study was to determine the incidence of and risk factors for drug resistance of Streptococcus pneumoniae, and its impact on the outcome among hospitalized patients of pneumococcal pneumonia acquired in the community. Consecutive patients with culture-proven pneumococcal pneumonia were prospectively studied with regard to the incidence of pneumococcal drug resistance, potential risk factors, and in-hospital outcome variables. A total of 101 patients were studied. Drug resistance to penicillin, cephalosporin, or a macrolide drug was found in pneumococci from 52 of the 101 (52%) patients; 49% of these isolates were resistant to penicillin (16% intermediate resistance, 33% high resistance), 31% to cephalosporin (22% intermediate and 9% high resistance), and 27% to a macrolide drug. In immunocompetent patients, age > 65 yr was significantly associated with resistance to cephalosporin (odds ratio [OR]: 5.0; 95% confidence interval [CI]: 1.3 to 18.8, p = 0.01), and with the presence of > 2 comorbidities with resistance to penicillin (OR: 4.7; 95% CI: 1.2 to 19.1; p < 0.05). In immunosuppressed patients, bacteremia was inversely associated with resistance to penicillin and cephalosporin (OR: 0.04; 95% CI: 0.003 to 0.45; p < 0.005; and OR: 0.46; 95% CI: 0.23 to 0.93; p < 0.05, respectively). Length of hospital stay, severity of pneumonia, and complications were not significantly affected by drug resistance. Mortality was 15% in patients with any drug resistance, as compared with 6% in those without resistance. However, any drug resistance was not significantly associated with death (relative risk [RR]: 2.5; 95% CI: 0.7 to 8.9; p = 0.14). Moreover, attributable mortality in the presence of discordant antimicrobial treatment was 12%, as compared with 10% (RR: 1.2; 95% CI: 0.3 to 5.3; p = 0.67) in the absence of such treatment. We conclude that the incidence of drug-resistant pneumococci was high. Risk factors for drug resistance included advanced age, comorbidity, and (inversely) bacteremia. Outcome was not significantly affected by drug resistance.

Streptococcus pneumoniae continues to be the leading agent of adult community-acquired pneumonia (CAP), accounting for approximately 30% of episodes of CAP in most series (1-3). More recently, human immunodeficiency virus (HIV)-infected patients have been identified as being at particular risk for pneumococcal bacteremia and pneumonia (4, 5). Mortality from pneumococcal pneumonia despite effective antimicrobial treatment has remained unchanged in the past decade, and is about 10% in nonbacteremic pneumonia and up to 30% in bacteremic pneumonia, in the elderly, and in severe pneumonia (1, 2, 6, 7).

The incidence of penicillin resistance in pneumococcal isolates is increasing worldwide. It has developed at different rates in various countries (8, 9). The highest resistance rates in Europe have been identified in France, Hungary, Spain, Portugal, and Iceland, reaching 50% of isolates (8-17). In the United States, resistance rates of more than 20% have been reported (18-20). Penicillin resistance tends to be associated with increasing resistance also to cephalosporins (21). Additionally resistance to macrolide antibiotics occurs at rates of less than 5% (e.g., in Germany) to more than 25% (e.g., in France) (10). The rate is currently about 10% in the United States (20). HIV infection has also been shown to represent a risk factor for infection with drug-resistant pneumococci (22).

Given the high incidence of both pneumococcal pneumonia and drug resistance, the eminent importance of studying the effect of resistance to penicillin and cephalosporin on mortality and other important outcome variables in this disease is obvious. To date, few studies have adressed this issue. For instance, a large study in Barcelona, Spain found no excess mortality from drug resistance in severe pneumococcal pneumonia after adjustment for other predictors of mortality (16). We therefore investigated the incidence of and risk factors for pneumococcal drug resistance, and its impact on outcome in community-acquired pneumococcal pneumonia.

Patient Population

All consecutive patients hospitalized at our institution (an 1,000-bed teaching hospital) with community-acquired pneumococcal pneumonia were prospectively studied during an 18-mo-period (from October 1, 1996, to March 31, 1998). Pneumonia was defined as the presence of symptoms of lower respiratory tract infection along with a new infiltrate on chest radiography and no emerging alternative diagnosis during follow-up. It was considered as acquired in the community in a patient not hospitalized for at least the previous 4 wk.

Data Collection

The following data from all patients were recorded at admission: age, gender, hospitalizations within the year preceding the study, smoking and alcohol habits, comorbid illnesses, antimicrobial treatment for pneumonia prior to hospital admission, clinical symptoms (body temperature, presence of chills, chest pain, cough, expectoration, dyspnea, confusion [i.e., disorientation with regard to person, place, or time that was not known to be chronic, stupor, or coma]), physical examination (i.e., presence of rales, respiratory rate, heart rate, and arterial systolic and diastolic blood pressure), results of blood gas analysis (PaO2 , PaCO2 , PaO2 /Fi O2 ), chest radiograph pattern (alveolar, interstitial, or mixed infiltrate, multilobar involvement, uni- versus bilateral involvement, pleural effusion), leukocyte count, and serum creatinine concentration. At the clinical endpoints of hospital discharge or death, the following parameters were additionally retrieved: results of microbial investigations including susceptibility testing, admission to the intensive care unit (ICU), Acute Physiology and Chronic Health Evaluation Score (APACHE II) at ICU admission, complications (i.e., pleural empyema, septic shock as defined by Bone and coworkers [23]), renal failure (serum creatinine ⩾ 2 mg/dl at any time during hospitalization in a patient without known previous impairment of renal function, an increase in serum-creatinine of ⩾ 2 mg/dl from known baseline values in patients with previous impairment of renal function, or acute renal failure requiring dialysis), requirement for mechanical ventilation, in-hospital antimicrobial treatment, and in-hospital outcome (days of hospital stay and survival or death within 30 d of in-hospital treatment).


Immunosuppression was defined as the presence of HIV infection, solid organ transplantation, or hematologic malignancy with known associated immune defects (in humoral or cellular immunity). Oral steroid treatment was considered immunosuppresive when it consisted of doses > 20 mg prednisone per day for > 2 wk. Smokers were defined as current smokers of ⩾ 10 cigarettes/d during at least the year preceding the study. Alcohol abuse was defined as estimated daily consumption of at least 80 g alcohol for at least the preceding year. Comorbidities were defined as follows:

Cardiac comorbid illness: treatment for coronary artery disease or congestive heart failure or presence of valvular heart disease.

Pulmonary disease: treatment for asthma or chronic obstructive pulmonary disease (COPD), or presence of interstitial lung disorders.

Renal disease: preexisting renal disease with documented abnormal serum creatinine levels outside the pneumonia episode.

Hepatic disease: preexisting viral or toxic hepatopathy.

Central nervous system disorders: symptomatic acute or chronic vascular or nonvascular encephalopathy, with or without dementia.

Diabetes mellitus: diagnosis of intolerance to glucose and treatment with oral antidiabetics or insulin.

Neoplastic illness: any solid tumor active at the time of presentation or requiring antineoplastic treatment within the preceding year.

Microbiologic Evaluation

Regular sampling was done of sputum, blood for two blood cultures, and specimens for paired serology (at admission and within the 4th and 8th weeks thereafter). Pleural puncture, transthoracic needle puncture, tracheobronchial aspiration, and flexible fiberopticbronchoscopy with a protected specimen brush (PSB) and bronchoalveolar lavage (BAL) as additional diagnostic techniques were used according to the clinical judgment of the physician in charge.

Sputum was Gram stained. Representative sputum originating from the lower respiratory tract was defined as that containing > 25 granulocytes and < 10 epithelial cells per low power field (lpf: total magnification: ×100) (24). Such validated sputum, pleural fluid, and transthoracic needle aspiration samples, as well as undiluted and serially diluted samples of tracheobronchial aspirates, PSB material, and BAL fluid (BALF) were plated on sheep-blood agar, chocolate agar, and Sabouraud's agar. In addition, blood cultures were produced as appropriate. Identification of microorganisms was done according to standard methods (25). Results of quantitative cultures were expressed as colony-forming units per ml (cfu/ml).

Susceptibility to antimicrobial agents was investigated with an automated microdilution system (Sensititre; AccuMed International Ltd., West Sussex, UK, for S. pneumoniae) according to the recommendations of the manufacturer, and was evaluated according to the guidelines established by the National Committee for Clinical Laboratory Standards (26). A pneumococcal isolate was considered susceptible to penicillin if the minimal inhibitory concentration (MIC) was ⩽ 0.06 μg/ml, to be of intermediate resistance if the MIC was 0.12 to 1 μg/ml, and to be highly resistant if the MIC was ⩾ 2 μg/ml. Intermediate resistance to cephalosporin was defined by an MIC of cefotaxime of 1 μg/ml, and high resistance was defined by an MIC of ⩾ 2 μg/ml. According to current convention in the United States, the definition of resistance to penicillin and cephalosporin for infections other than otitis or meningitis is an MIC ⩾ 2 μg/ml (corresponding to high resistance according to the conventional definition). Resistance to macrolide antibiotics was defined by an MIC of ⩾ 1 μg/L. Susceptibility to imipenem was considered in the presence of an MIC of ⩽ 0.12 μg/ml, intermediate resistance in the presence of an MIC of 0.25 to 0.5 μg/ml, and high resistance in the case of an MIC of ⩾ 1 μg/L. Susceptibility to ciprofloxacin was considered in the presence of an MIC of ⩽ 1 μg/ ml, intermediate resistance in the presence of an MIC of 2 μg/ml, and high resistance in the presence of an MIC of ⩾ 4 μg/ml.

Diagnostic Criteria

The diagnosis of pneumococcal pneumonia was considered probable in patients with a sputum culture positive for S. pneumoniae. It was classified as definite if one of the following criteria was met: a culture yielding S. pneumoniae exclusively, or presence of the organism in: (1) blood culture; (2) pleural fluid; (3) a transthoracic needle aspirate; (4) a tracheobronchial aspirate with ⩾ 105 cfu/ml; (5) a PSB sample with ⩾ 103 cfu/ml; or (6) a BALF specimen with ⩾ 104 cfu/ml.

Antimicrobial Treatment

A general guideline for treatment consisted of considering an initial antimicrobial regimen including a third-generation cephalosporin and a macrolide. The exact design of the initial antimicrobial treatment regimen was the responsibility of the physician in charge. Adjustments of antimicrobial treatment were made as soon as the results of susceptibility testing were available. Generally, cephalosporin treatment in the presence of intermediate resistance to cephalosporins was modified only in case of clinical nonresponse to this treatment.


Results are expressed as means ± SD. Continuous variables were compared through Student's t test, categorical variables were compared through the chi-square test or Fisher's exact test where appropriate. Associations of drug resistance with categorical outcome variables were assessed with the chi-square test, giving odds ratios (ORs) and relative risks (RRs), respectively. Multivariate analysis of factors associated with mortality was done by stepwise forward logistic regression. All reported p values are two-tailed. The level of significance was set at 5%.

Patient Descriptors

Overall, 101 patients with probable and definite pneumococcal pneumonia acquired in the community were included. Five patients were admitted from a nursing home. The study population comprised 79 immunocompetent (56 male and 23 female) and 22 immunosuppressed (17 male and five female) patients. Immunosuppression included HIV-infection (n = 18), chronic lymphoblastic leukemia, bone marrow transplantation, renal transplantation, or COPD with chronic steroid treatment in a dose of 30 mg prednisone-equivalent per day. The mean ages of the immunocompetent and immunosuppressed patients were 67 ± 15 yr and 38 ± 12 yr, respectively (p < 0.0001).

Comorbid illnesses are listed in Table 1. Overall, 44 patients had one, 19 had two, nine had three, two had four, and one had five comorbid illnesses. Immunocompetent patients more frequently had comorbidities (79% versus 48%, p < 0.01). The main clinical and radiographic characteristics are summarized in Table 1. There were no significant differences between the immunocompetent and immunosuppressed groups.


Comorbid illnesses
 Cardiac 8
 Renal 5
 CNS 5
 Diabetes mellitus24
 Miscellaneous 2
Clinical symptoms
 Chest pain45
 Body temperature ⩾ 38.3° C57
Chest radiography
 Alveolar infiltrates77
 Interstitial infiltrates 3
 Mixed interstitial infiltrates19
 Bilateral involvement19
 Multilobar involvement (> 2 lobes)21
 Pleural efffusion15

Diagnosis of Pneumococcal Pneumonia

The diagnosis of pneumococcal pneumonia was based on validated sputum in 42 patients. A further 12 patients with positive sputum samples also had positive blood cultures. Overall, 47 patients had bacteremia. The incidence of bacteremia was not different in immunocompetent and immunosupressed patients (37 of 79 [47%] versus 10 of 22 [46%], respectively; p = NS). Overall, six patients with severe pneumonia who were admitted to the ICU (23%) had a diagnosis of pneumococcal pneumonia made exclusively through validated sputum. Blood cultures had a yield of 55% (47 of 85). The diagnostic results are summarized in Table 2.


Medium of DiagnosisNo. of PatientsNo. of Samples/ No. (%) Positive
Validated sputum4269/54 (78)
Blood culture3185/47 (55)
Pleural fluid 2  9/5 (56)
T TBAS 913/12 (92)
P PSB  7/1 (14)
T TTP  4/2 (50)
Validated sputum + blood culture12
Pleural fluid + blood culture 1
TBAS + blood culture 1
Pleural fluid + blood culture + TTP 1
Pleural fluid + blood culture + TBAS + TTP 1

Definition of abbreviations: PBS = protected specimen brush; TBAS = tracheobronchial aspirate; TTP = transthoracic puncture.


Coinfections were present in 24 patients (24%), in whom 26 pathogens were identified. These included Hemophilus influenzae (n = 9; five validated sputum, three TBAS, one PSB), Moraxella catarrhalis (n = 2; both validated sputum), Pseudomonas aeruginosa (n = 2; both validated sputum), Chlamydia pneumoniae (n = 2; both through a fourfold rise in antibody titer [i.e., seroconversion]), Mycoplasma pneumoniae (n = 1; seroconversion), Coxiella burnetii (n = 1; seroconversion), Legionella sp. (n = 2; single titre ⩾ 1:128 and seroconversion), Influenza A (n = 4; seroconversion), Parainfluenzavirus 1 (n = 2; seroconversion), and respiratory syncytial virus (n = 4; seroconversion). Coinfections were significantly more frequent in immunocompetent patients (29% versus 5%, p < 0.05).

Antimicrobial Susceptibility Patterns

The overall resistance rate to penicillin, cephalosporin, or macrolide was 52 (52%). Specifically, 49% of these isolates were resistant to penicillin (16% intermediate and 33% high resistance), 31% to cephalosporin (22% intermediate and 9% high resistance), and 27% to macrolide. Resistance to ciprofloxacin was 17% (14% intermediate and 3% high resistance). Resistance to imipenem was 25% (all intermediate). All isolated strains were susceptible to vancomycin.

Three strains were exclusively resistant to macrolide. Of 49 strains resistant to penicillin, 42 (86%) were susceptible to ciprofloxacin (including 31 of 33 [94%] that were highly resistant to penicillin), twenty nine of 31 (94%) strains were resistant to cephalosporin (including eight of nine [89%] that were highly resistant to cephalosporin), and 23 of 27 (85%) strains were resistant to erythromycin. Conversely, 10 of 52 (19%) strains susceptible to penicillin were resistant to ciprofloxacin, 15 of 70 (21%) strains were susceptible to cephalosporin, and 13 of 74 (18%) strains were susceptible to erythromycin. All nine strains highly resistant to penicillin and cephalosporin also showed intermediate resistance to imipenem.

Susceptibility patterns to penicillin, cephalosporin, macrolide drugs, ciprofloxacin, and vancomycin, as well as multidrug resistances stratified according to immunocompetence, are given in Table 3.


Immunocompetent Patients (n = 79)Immunosuppressed Patients (n = 22)
n (%)NS (%)n (%)NS
 Susceptible44 (56)3 (7) 8 (36)0
 Intermediately resistant11 (14)1 (10) 5 (23)1 (20)
 Highly resistant24 (30)5 (21) 9 (41)1 (11)
 Susceptible56 (71)4 (7)14 (64)1 (7)
 Intermediately resistant15 (19)3 (20) 7 (32)1 (14)
 Highly resistant 8 (10)2 (25)1 (5)0
 Susceptible58 (73)8 (14)16 (73)2 (13)
 Resistant21 (27)1 (5) 6 (27)1 (17)
 Susceptible66 (84)9 (14)18 (82)2 (11)
 Intermediately resistant11 (14)0 3 (14)0
 Highly resistant2 (2)01 (4)0
 Susceptible61 (77)5 (8)15 (68)1 (7)
 Intermediately resistant18 (23)4 (22) 7 (32)1 (14)
 Susceptible79 (100)9 (11)22 (100)2 (9)
 Highly penicillin and   (intermediately and highly)   cephalosporin resistant23 (29)5 (22) 8 (36)1 (13)
 Highly penicillin and   highly cephalosporin resistant 8 (10)2 (25)1 (5)0
 Highly penicillin and   erythromycin resistant11 (14)1 (9) 5 (23)1 (20)
 Highly penicillin and   highly cephalosporin and   erythromycin resistant1 (1)01 (5)0

Definition of abbreviation: NS = nonsurvivors.

*For definition of immunosuppression see paragraph on Patient Descriptors. p = NS for all comparisons of proportions of nonsurvivors.

In isolates from blood cultures, 38% were penicillin resistant (18 of 47, including five isolates that showed intermediate resistance and 13 that showed high resistance) and 28% were cephalosporin resistant (13 of 47, including nine isolates that showed intermediate and four that showed high resistance).

Risk Factors for Penicillin and Cephalosporin Resistance

Potential factors that were tested for an association with penicillin and cephalosporin resistance included age > 65 yr, sex, hospitalization within the prior year, cigarette smoking, alcoholism, single comorbidities, more than two comorbidities, HIV infection, immunosuppression, ambulatory antimicrobial pretreatment, mechanical ventilation, and invasive disease. Since these factors were not significantly different in patients with isolates showing intermediate as opposed to high resistance to penicillin and cephalosporin, analysis of risk factors associated with penicillin and cephalosporin resistance was done without regard to the level of resistance. Conversely, the analysis of risk factors for drug resistance was extended to an analysis of both immunocompetent and immunosuppressed patients.

With regard to the whole study population, there was a clear albeit nonsignificant trend toward a higher incidence of penicillin and cephalosporin resistance in immunosuppressed patients (33% versus 64% and 29% versus 37%, respectively). Moreover, there was a trend for bacteremia to be inversely associated with penicillin resistance (OR: 0.46; 95% CI: 0.2 to 1.02; p = 0.06). No other factors tested were associated with resistance to either drug.

In the immunocompetent group, age > 65 yr was significantly associated with resistance to cephalosporin (OR: 5.0; 95% CI: 1.3 to 18.8; p = 0.01). Penicillin resistance was associated with the presence of more than two comorbidities (OR: 4.7; 95% CI: 1.2 to 19.1; p < 0.05). In the immunosuppressed group, bacteremia was inversely associated with penicillin as well as with cephalosporin resistance (OR: 0.04; 95% CI: 0.003 to 0.45; p < 0.005, and OR: 0.46; 95% CI: 0.23 to 0.93; p < 0.05, respectively).

Antimicrobial Treatment Regimen

Overall, 10 patients had ambulatory antimicrobial pretreatment (oral amoxycillin [n = 2], macrolide [n = 3], cephalosporin [n = 2], and quinolone, cotrimoxazole, and cefixime [n = 1 each]). Initial in-hospital antimicrobial treatment consisted of: a macrolide plus a third-generation cephalosporin (n = 62) or cefepime (n = 1); a macrolide plus a third-generation cephalosporin plus a third agent (n = 11; aminoglycoside [n = 4], cloxacillin, cotrimoxazole [n = 2 each], clindamycin, rifampicin, vancomycin [n = 1 each]); a third-generation cephalosporin alone (n = 13); a third-generation cephalosporin plus a second nonmacrolide agent (n = 7; aminoglycoside [n = 2], aminopenicillin plus a β-lactamase inhibitor, cloxacillin, clindamycin, rifampicin, cotrimoxazole [n = 1 each]); a macrolide plus one or two noncephalosporin agents (n = 3; quinolone, vancomycin, aminoglycoside plus cotrimoxazole [n = 1 each]); and miscellaneous monotherapies (n = 4; aminopenicillin plus a β-lactamase inhibitor [n = 2], imipenem and cefepim [n = 1 each]). Third-generation cephalosporins were always administered intravenously. Patients with severe pneumonia or drug resistance who received third-generation cephalosporins were treated with ceftriaxone in a dose of 2 g/d or with cefotaxime in a dose of 2 g given thrice daily.

Initial antimicrobial treatment was changed in 20 of 52 patients (39%) with drug-resistant pneumococci. These changes included simplifications to monotherapy (n = 11), the addition of one further drug (n = 1), the substitution of one or more drugs (n = 5), and a complete change of the antimicrobial regimen (n = 3).

Antimicrobial treatment of 52 patients with drug-resistant strains of S. pneumoniae is summarized in Table 4. Overall, 17 patients received antimicrobial treatment exclusively with antimicrobial agents to which pneumococcal strains were resistant. These included 16 patients treated with ceftriaxone or cefotaxime in the presence of strains showing intermediate (n = 15) or high resistance (n = 1) to cephalosporins, and one patient initially treated with aminopencillin plus a β-lactamase inhibitor, followed by a third-generation cephalosporin in the presence of high resistance to penicillin and intermediate resistance to cephalosporin.


Antimicrobial Susceptibility PatternInitial Antimicrobial TreatmentCases/Deaths
Pi/Cs/Ms, n = 8Macrolide + third-generation C3/0
Macrolide + vancomycin + imipenem1/1
Macrolide + aminoglycoside + cefepim1/1
Third-generation C + cloxacillin1/0
Pr/Cs/Ms, n = 2Macrolide + third-generation C1/0
Third-generation C1/0
Pr/Ci/Ms, n = 8Macrolide + third-generation C4/1
Third-generation C3/1*
Third-generation C + aminoglycoside1/0*
Pr/Cr/Ms, n = 7Macrolide + third-generation C5/1
Third-generation C + clindamycin/vancomycin1/1
Pi/Cs/Mr, n = 8Third-generation C + macrolide5/0
Third-generation C + macrolide + cotrimoxazole1/0
Third-generation C1/0
Quinolone + macrolide1/0
Pr/Ci/Mr, n = 14Third-generation C + macrolide8/2*
Third-generation C + macrolide + aminoglycoside1/0*
Third-generation C2/0*
Third-generation C + cotrimoxazole1/0
Third-generation C + rifampicin1/0
Aminopenicillin plus β-lactamase-inhibitor/third-generation C1/0*
Pr/Cr/Mr, n = 2Third-generation C + macrolide1/0*
Third-generation C + macrolide + cloxacillin/vancomycin1/0
Ps/Cs/Mr, n = 3Third-generation C + macrolide2/0
Third-generation C + macrolide + aminoglykoside1/0

Definition of abbreviations: C = cephalosporin; i = intermediate resistance; M = macrolide; P = penicillin; r = highly resistant; s = susceptible.

*Patients treated exclusively with antimicrobial agents to which pneumococcal strains were not fully susceptible or resistant.

Outcome and Prognostic Analysis

The mean duration of hospitalization was 11 ± 8 d (range: 2 to 47 d) (11 ± 8 d in immunocompetent and 9 ± 9 d in immunosuppressed patients, p = NS). Overall, 26 patients were admitted to the ICU (n = 21 [27%] immunocompetent and n = 5 [23%] immunosuppressed patients [p = NS). Mortality was 11 of 101 (11%; n = 9 [11%] among immunocompetent and n = 2 [9%] among immunosuppressed patients [p = NS]).

Hospital stay (11 ± 9 d versus 10 ± 7 d, and 11 ± 8 d versus 11 ± 7 d), empyema (6% versus 4%, and 6% and 3% of patients), ICU admission (25% versus 27%, and 27% versus 23% of patients), mean APACHE II score at ICU admission (24 ± 11 versus 19 ± 10, and 23 ± 11 versus 18 ± 9), septic shock (18% versus 14%, and 17% versus 13% of patients), renal failure (16% versus 19%, and 23% versus 16% of patients), and the requirement for mechanical ventilation (20% versus 15%, and 20% versus 13% of patients) were not significantly different in patients without and with penicillin and cephalosporin resistance, respectively. The same was true when considering only high resistance to both drugs.

Three patients with susceptible strains of S. pneumoniae died of pneumonia after Days 5, 9, and 10 of adequate antimicrobial treatment. An overview of eight deaths of patients with drug resistant strains is given in Table 5. Only three cases (IDNR 2, 4, and 8) received discordant antimicrobial treatment in the presence of drug resistance (crude mortality with discordant antimicrobial treatment was 18% [three of 17 patients treated only with drugs to which pneumococcal isolates were resistant]). Death was attributable to pneumonia in only two of these cases (attributable mortality with discordant antimicrobial treatment 12% [two of 17 patients treated only with drugs to which isolates were resistant]). In both cases (IDNR 2 and 8), results of susceptibility testing were not available while the patients were alive. These cases may be regarded as instances of true failure of antimicrobial treatment with cephalosporins in the presence of intermediate resistance to cephalosporin. However, additional risk factors for death were HIV infection (IDNR 2), and advanced age and diabetes mellitus as well as breast carcinoma (IDNR 8).


IDNRAge (yr)Type of Antimicrobial ResistanceAntimicrobial Treatment AdministeredDuration of Treatment (d )Attributable Mortality/Further Complications
155Pi/Cs/MsM + VC + IP + T/(C)26Yes/Empysema suspected superinfection
231Ph/Ci/MrC + M/(T) 2Yes
382Ph/Ch/MsC + M10Yes
474Ph/Ci/MrC + M15No/Coinfection with H. influenzae death during dialysis
589Ph/Ch/MsC + Cl*/(CX + VC)16Yes/Empyema heart failure
676Ph/Ci/MsC + M23No/Severe GI bleeding
757Pi/Cs/MsM + T + CP 2Yes
891Ph/Ci/MsC 5Yes

Definition of abbreviations: C = ceftriaxone; Cl = clindamycin; CP = cefepim; CX = cefotaxime; h = highly resistant; i = intermediate resistance; IDNR = identification number; IP = imipenem; M = macrolide; s = susceptible; T = tobramycin; VC = vancomycin.

*Previous ambulatory treatment with macrolide.

Factors significantly associated with mortality in univariate analysis included systolic blood pressure < 90 mm Hg and confusion at admission, as well as septic shock, renal failure, and requirement for mechanical ventilation at any time. Conversely, age > 65 yr, the presence of comorbid illness, HIV infection, coinfection, respiratory rate > 30 breaths/min, PaO2 /Fi O2 < 250, and multilobar involvement were not significantly associated with mortality. In multivariate analysis including the five variables significantly associated with death, confusion and renal failure proved to be independently associated with death (Table 6).


CriteriaIncidence/DeathsRelative Risk95% CIp Value
 Clinical criteria at admission
  Systolic blood pressure < 90 mm Hg 7/3 5.01.7–14.90.03
  Confusion26/8 7.72.2–26.8 0.0007
 Clinical criteria at admission or during    follow-up
  Septic shock16/814.24.2–47.7< 0.0001
  Renal failure18/920.84.9–88.0< 0.0001
  Mechanical ventilation18/7 8.12.6–24.7 0.0004
 Antimicrobial susceptibility
  Resistance to penicillin    (intermediate and high)49/8 2.80.8–10.10.09
  Resistance to cephalosporin    (intermediate and high)31/6 2.70.9–8.20.09
  Resistance to penicillin and cephalo-    sporin (intermediate and high)31/6 2.70.9–8.20.09
  Resistance to penicillin (high)33/6 2.50.8–7.50.17
  Resistance to cephalosporin (high) 9/2 2.30.6–8.90.26
  Resistance to penicillin and    cephalosporin (high) 9/2 2.30.6–8.90.26
  Resistance to penicillin,    cephalosporin, or macrolide    (intermediate and high)52/8 2.50.7–8.90.14
 Discordant antimicrobial treatment
  Crude mortality17/3 1.90.6–6.30.39
  Attributable mortality17/2 1.20.3–5.30.67
 Confusion 5.91.1–31.60.036
 Renal failure28.34.9–162.2 0.0002

Definition of abbreviation: CI = confidence interval.

Mortality was 15% (eight of 52) versus 6% (three of 49) among patients with any antimicrobial resistance as compared with those without resistance, and 16% (eight of 49) versus 6% (three of 52) and 19% (six of 31) versus 7% (five of 70) among patients with intermediate or high resistance to penicillin and cephalosporin, respectively. Mortality was highest in patients with highly penicillin-, cephalosporin-, and both penicillin- and cephalosporin-resistant strains (18% [six of 33], 22% [two of 9], and 22% [two of 9], respectively). These differences were not significant. Moreover, discordant antimicrobial treatment was not significantly associated with death (18% [three of 17] and 12% [two of 17] versus 10% [eight of 84] crude and attributable mortality, respectively). Accordingly, the additional adjustment for penicillin or cephalosporin resistance did not change the results of the multivariate analysis of independent factors associated with death (Table 6).

The most important findings of this study were that: (1) the incidence of resistance to penicillin, cephalosporin, and macrolide was very high (49%, 31%, and 27%, respectively, with an incidence of 33% and 9% of high resistance to pencillin and cephalosporin, respectively); (2) drug resistance was associated with age > 65 yr and more than two comorbidities in the immunocompetent group, and was inversely associated with invasive disease in the immunosuppressed group; (3) important outcome variables were not affected by the presence of drug resistance. Attributable mortality in the presence of discordant antimicrobial treatment was virtually identical to that with concordant antimicrobial treatment (12% versus 10%).

In contrast to a previous study in Barcelona (16), we included patients with pneumococcal pneumonia proven by sputum culture validated by strict standard criteria. Since penicillin resistance has been shown to be independently associated with noninvasive pneumococcal infection (14, 15, 17), which was also confirmed in our study, we thought it would be important not to discard the subgroup of patients with noninvasive infection but not necessarily less severe pneumonia. In our study, 23% of patients with severe pneumonia requiring ICU admission had a diagnosis of pneumococcal pneumonia made exclusively from validated sputum. Nevertheless, the incidence of bacteremia was high (47%), most probably because of a low rate of ambulatory antimicrobial pretreatment (10%). It should be stressed that our series covered a rather short period of only 18 mo, including two winter/spring seasons without evidence of outbreaks, and did not include nosocomial pneumococcal pneumonia. In addition, we controlled for the presence of immunosuppression in all of our comparisons.

Our results confirm the high rates of drug resistance in our region. In fact, these rates are significantly higher than those reported from another hospital in Barcelona for the period between 1989 and 1993 (35% resistance to penicillin, including 15% high resistances, 9% resistance to cephalosporin, and 13% resistance to erythromycin) (16). In the most recent epidemiologic survey of pneumococcal respiratory infections in Spain, including 14 centers in different regions of the country, the overall resistance rate to penicillin was 36.5%, to cefotaxime and ceftriaxone 15.1% and 9.0%, respectively, and to macrolides 34.6% (28). Thus, our findings are representative of the current Spanish epidemiologic situation with regard to drug-resistant pneumococcal pneumonia. Somewhat intriguingly, we found that about two-thirds of penicillin-resistant strains of pneumococcus were highly resistant, which contrasts with two previous reports from Spain of a proportion of highly resistant strains of about 50% (16, 17). We found that high resistance to penicillin and cephalosporin was significantly associated with age ⩾ 65 yr in immunocompetent patients, confirming previous findings (17). In addition, as expected, there was a trend for penicillin and cephalosporin resistance to occur more frequently in immunosuppressed patients (14, 29).

We also found an inverse association of invasive disease with penicillin resistance, which was again in concordance with the findings of others (14, 15, 17), although the association was significant only in the immunosuppressed group. In this group, this inverse association was also found for cephalosporin resistance, a finding that has not been reported previously. However, given the high frequency of penicillin- and cephalosporin-resistant isolates from blood culture (38% and 28%, respectively), this finding should be interpreted with caution. In the immunocompetent group, age > 65 yr and multimorbidity were identified as risk factors for drug resistance, in accord with the results in previous reports (14, 16, 17). Because we were unable to reliably assess in every patient the history of β-lactam treatment, which has been found to be an independent predictor of drug resistance by others (17, 29), we tested only the impact of hospitalization within the year preceding our study and that of recent antimicrobial pretreatment with β-lactam drugs prior to hospital admission. Both failed to be associated with drug resistance. Nevertheless, previous repetitive use of β-lactam antibiotics most probably does represent an additional risk factor of resistance.

Overall mortality in our study was 11%, which is in the expected range (1, 2). Neither the duration of hospital stay nor the proportion of patients with pneumonia-related complications and severe pneumonia were significantly different among patients with and without drug resistance. Although mortality appeared to be threefold higher in patients with any drug resistance (18% versus 6%), this difference was not significant. Likewise, drug resistance was not an independent predictor of death after adjustment for other factors associated with mortality in a multivariate model. Thus, the apparently higher mortality rates of patients with drug resistance were most probably due to other concomitant risk factors for mortality, such as advanced age and multimorbidity. Further evidence that drug resistance did not independently affect mortality can be derived by comparing patients with concordant versus discordant treatment (drugs to which the organism was resistant). The relative risk (RR) for attributable mortality in the presence of discordant antimicrobial treatment was near 1, indicating that there was no excess mortality due to seemingly inappropriate treatment.

The majority of patients were initially treated with a combination of a third-generation cephalosporin and a macrolide. Only 17 patients received treatment that did not include at least one agent to which the pneumococcal organism was fully susceptible. Of these 17 patients, only two died from acute pneumonia. Cephalosporin resistance was intermediate in both cases. Although these deaths may be attributed to true antimicrobial treatment failure, the effect of resistance remains debatable, since other risk factors for death were present. It has been estimated that the breakpoints for the failure of intravenous β-lactam treatment are likely to be in the region of ⩾ 8 μg/ml (30). However, resistance levels of this magnitude are not usually encountered with penicillin, ceftriaxone and cefotaxime, and carbapenems, and we did not encounter them. The 15 patients successfully treated with ceftriaxone or cefotaxime despite intermediate or high cephalosporin resistance (in the absence of another effective agent to which pneumococcal organisms were susceptible) are additional evidence that the concentrations of cephalosporins achieved at the site of infection, at least when these drugs were given intravenously, were sufficient to overcome the corresponding degrees of resistance of the particular strains.

With regard to second-line drugs, all strains were susceptible to vancomycin but 25% had intermediate resistance to imipenem, and 17% had intermediate to high resistance to ciprofloxacin. Since all nine strains highly resistant to penicillin and cephalosporin were also intermediately resistant to imipenem, this drug does not represent an ideal alternative choice. Conversely, susceptibility to ciprofloxacin was not predictable from susceptibility patterns to other drugs. Therefore, in selected patients with pneumococci having high resistance to penicillin, a cephalosporin, and/or a macrolide, ciprofloxacin may be an option. However, newer quinolones, such as levofloxacin, grepafloxacin, and trovafloxacin, which were not tested in our study, have higher in vitro activities against S. pneumoniae and are likely to gain priority in drug-resistant pneumococcal pneumonia.

In conclusion, we found that drug resistance to S. pneumoniae in patients with CAP treated with an initial antimicrobial regimen designed with an expected high prevalence of drug resistance was not associated with an impaired clinical outcome in terms of morbidity and mortality. Nevertheless, the high and still increasing incidence of drug resistance to pneumococcal organisms is still a matter of concern. Continuous surveillance of patterns of pneumococcal resistance, together with judicious use of effective antimicrobial treatment, is mandatory in order to prevent future disease consequences of resistance.

Supported by grant SGR 00086 from the Commisionat per a Universitats i Recerca de la Generalitat de Catalunya 1997 and by the Fundació Clı́nic / CIRIT, IDIBAPS Hospital Clı́nic Barcelona.

1. British Thoracic Society and the Public Health Laboratory ServiceCommunity-acquired pneumonia in adults in British hospitals in 1982– 1983: a survey of aetiology, mortality, prognostic factors and outcome. Q. J. Med.2391987195220
2. Örtquist A., Hedlund J., Grillner L., Jalonen E., Kallings I., Leionen M., Kalin M.Aetiology, outcome and prognostic factors in patients with community acquired pneumonia requiring hospitalisation. Eur. Respir. J.3199011051113
3. Fang G. D., Fine M., Orloff J., Arisumi D., Yu V. L., Kapoor W., Grayston J. T., Wang S. P., Kohler R., Muder R. R., Yee Y. C., Rihs J. D., Vickers R. M.New and emerging etiologies for community-acquired pneumonia with implications for therapy: a prospective multicenter study of 359 cases. Medicine691990307316
4. Redd S. C., Rutherford G. W. I., Sande M. A., Lifson A. R., Hadley W. K., Facklam R. R., Spika J. S.The role of human immunodeficiency virus infection in pneumococcal bacteremia in San Francisco residents. J. Infect. Dis.162199010121017
5. Burack J. H., Hahn J. A., Saint-Maurice D., Jakobson M.Microbiology of community-acquired bacterial pneumonia in persons with and at risk for human immunodeficiency virus type 1 infection. Am. J. Med.154199425892596
6. LaCroix A. Z., Lipson S., Miles T. P., White L.Prospective study of pneumonia hospitalizations and mortality of US older people: role of chronic conditions, health behaviours, and nutritional status. Public Health Rep.1041989350360
7. Moine P., Vercken J. B., Chevret S., Gajdos P.the French Study Group of Community-Acquired Pneumonia in ICUSevere community-acquired pneumonia. Scand. J. Infect. Dis.271995201206
8. Appelbaum P. C.Antimicrobial resistance in Streptococcus pneumoniae: an overview. Clin. Infect. Dis.1519927783
9. Baquero F.Pneumococcal resistance to beta-lactam antibiotics: a global geographic overview. Microb. Drug Resist.11995115120
10. Geslin P., Buu-Hoi A., Fremaux A., Acar J. F.Antimicrobial resistance in Streptococcus pneumoniae: an epidemioogical survey in France 1970–1990. Clin. Infect. Dis.1519929598
11. Bedos J. P., 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.2219966372
12. Pallares R., Gudiol F., Linares J., Ariza J., Rufi G., Murgui L., Dorca J., Vildadrich P. F.Risk factors and response to antibiotic therapy in adults with bacterial pneumonia caused by penicillin-resistant pneumococci. N. Engl. J. Med.31719871822
13. Fenoll A., Bourgon C. M., Munoz R., Vicisio D., Casal J.Serotype distribution and antimicrobial resistance of Streptococcus pneumoniae isolates causing systemic infections in Spain, 1979–1989. Rev. Infect. Dis.1319915660
14. Garcia-Leoni M. E., Cercenado E., Rodeno P., Bernaldo de Quiros J. C. L., Martin-Hernandez D., Bouza E.Susceptibility of Streptococcus pneumoniae to penicillin: a prospective microbiological and clinical study. Clin. Infect. Dis.141992427435
15. Linares J., Pallares R., Alonso T., Perez J. L., Ayats J., Gudiol F., Viladrich P. F., Martin R.Trends in antimicrobial resistance of clinical isolates of Streptococcus pneumoniae in Bellvitge Hospital, Barcelona, Spain (1979–1990). Clin. Infect. Dis.15199299105
16. Pallares R., Linares J., Vadillo M., Cabellos C., Manresa F., Viladrich P. F., Martin R., Gudiol F.Resistance to penicillin and cephalosporin and mortality from severe pneumococcal pneumonia in Barcelona, Spain. N. Engl. J. Med.3331995474480
17. Clavo-Sanchez A. J., Giron-Gonzalez J. A., Lopez-Prieto D., Canueto-Quintero J., Sanchez-Porto A., Vergara-Campos A., Marin P., Casanova, Cordoba-Dona J. A.Multivariate analysis of risk factors for infection due to penicllin-resistant and multidrug-resistant Streptococcus pneumoniae: a multicenter study. Clin. Infect. Dis.24199710521059
18. Spika J. S., Facklam R. R., Pilkaytis B. D., Oxtoby M. J.Antimicrobial resistance of Streptococcus pneumoniae in the United States, 1979–1987. J. Infect. Dis.163199112731278
19. Breiman R. F., Butler J. C., Tenover F. C., Elliot J. A., Facklam R. R.Emergence of drug-resistant pneumococcal infections in the United States. J.A.M.A.271199418311835
20. Doern G. V., Brueggemann A., Holley H. P., Rauch A. M.Antimicrobial resistance of Streptococcus pneumoniae recovered from outpatients in the United States during the winter months of 1994 to 1995: results of a 30-center national surveillance study. Antimicrob. Agents Chemother.40199612081213
21. Thornsberry, C., S. D. Brown, Y. C. Yee, S. K. Bouchillon, J. K. Marler, and T. Rich. 1993. Increasing penicillin resistance in Streptococcus pneumoniae in the U.S.: effect on susceptibility to oral cephalosporins. Infect. Med. 10(D Suppl.):15–24.
22. Meynard J. L., Barbut F., Blum L., Guiget M., Chouaid C., Meyohas M. C., Picard O., Petit J. C., Frottier J.Risk factors for isolation of Streptococcus pneumoniae with decreased susceptibility to penicillin G from patients infected with human immunodeficiency virus. Clin. Infect. Dis.221996437440
23. Bone R. C., Balk R. A., Cerra F. B., Dellinger R. P., Fein A. M., Knaus W. A., Schein R. M., Sibbald W. J.Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest101199216441655
24. Murray T. J., Washington J. A.Microscopic and bacteriologic analysis of expectorated sputum. Mayo Clin. Proc.501975339344
25. Balows, A., and W. J. Harsier, Jr. 1991. Manual of Clinical Microbiology, 5th ed., section III. American Society of Microbiology, Washington, DC. 209–553.
26. National Committee for Clinical Laboratory Standards. 1997. Methods for Dilution Antimicrobial Susceptibility Test for Bacteria that Grow Aerobically, 4th ed. National Committee for Clinical Laboratory Standards, Wayne, PA. M7-A4.
27. Fine M. J., Auble T. E., Yealy D. M., Hanusa B. H., Weissfeld L. A., Singer D. E., Coley C. M., Marrie T. J., Kapoor W. N.A prediction rule to identify low-risk patients with community-acquired pneumonia. N. Engl. J. Med.3361997243250
28. Garcia de Lomas y grupo Espanol para vigilancia de patogenos respiratorios. 1998. Situacion epidemiologica actual y resistencia de los patogenos respiratorios en Espana. Med. Clin. (Barc.) 110(Suppl. 1):44–51.
29. Nava J. M., Bella F., Garau J., Lite J., Morera M., Marti C.Predictive factors for invasive disease due to penicillin-resistant Streptococcus pneumoniae: a population-based study. Clin. Infect. Dis.191994884890
30. Klugman K. P., Feldman C.The clinical relevance of antibiotic resistance in the management of pneumococcal pneumonia. Clin. Pulm. Med.41997190193
Correspondence and requests for reprints should be addressed to Dr. Antoni Torres, Hospital Clinic i Provincial, Servei de Pneumologia i Al·lergia Respiratoria, Villaroel 170, 08036 Barcelona, Spain. E-mail:

Dr. Ewig was a 1997 research fellow of the Medizinische Universitätsklinik and Poliklinik Bonn, Bonn, Germany.

Dr. Ruiz was a 1997 European Respiratory Society (ERS) research fellow at the Hospital Clı́nico de la Universidad de Chile, Santiago de Chile, Chile


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record