ACP medicine, 3rd Edition

Infectious Disease

Infections Due to Gram-Positive Cocci

Dennis L. Stevens PH.D., M.D., F.A.C.P.1

1Professor, University of Washington School of Medicine, Chief, Infectious Disease Section, Veterans Affairs Medical Center, Boise, Idaho

The author has no commercial relationships with manufacturers of products or providers of services discussed in this chapter.

September 2004


Pneumococci are gram-positive, lancet-shaped diplococci that may also grow in short chains. When cultured on blood agar plates, they form small, colorless mucoid colonies with characteristic central depressions. The colonies are surrounded by greenish discoloration of the blood agar, termed α-hemolysis. Laboratories have traditionally identified pneumococci on the basis of sensitivity to surface-active agents such as bile or optochin; rapid speciation can now be accomplished with a latex agglutination test or a DNA probe.

Although there is only one species of Streptococcus pneumoniae, there are at least 84 distinct serotypes, which are differentiated by the composition of the polysaccharide polymer that forms their outer capsule. Each capsular type is chemically and antigenically unique. The pneumococcal polysaccharide capsule is crucial to virulence. The capsule allows the bacteria to resist phagocytosis by leukocytes unless the organisms have been opsonized by antibody or serum complement components. Antibodies to capsular polysaccharide are essential for recovery from untreated pneumococcal pneumonia. Although such antibodies provide long-lasting immunity, the immunity is strictly type specific. Certain pneumococcal serotypes, such as the heavily encapsulated type 3 pneumococcus, are particularly virulent. In fact, only 23 serotypes account for about 80% of bacteremic pneumococcal infections in the United States; capsular polysaccharides from these 23 serotypes are incorporated into the polyvalent pneumococcal vaccine that has been available since 1983.

Although the polysaccharide capsule is the critical factor in determining the virulence of pneumococci, several proteins also contribute to the pathogenesis of pneumococcal infections. Surface protein A, neuraminidase, pneumolysin, and IgA protease appear to be the most important of these proteins.1


The nasopharynx is the natural habitat of the pneumococcus; humans are the only known hosts. The nasopharyngeal carrier rate varies widely, from a low of 5% to 10% to a high of up to 60% in closed populations during the winter.2 Pneumococci are fastidious, nonsporulating bacteria. Because they are rapidly killed by drying or by extremes in temperature, person-to-person spread by droplets requires close contact.

The pneumococcus remains the most important cause of bacterial pneumonia; it may account for as many as 500,000 cases each year in the United States. Like other respiratory tract infections, pneumococcal pneumonia is more common in winter. Cigarette smoking is the strongest independent risk factor for invasive pneumococcal disease in immunocompetent, nonelderly adults.3 Extremes of age and chronic disease or immunosuppression are also risk factors [see Table 1].4 Diseases that increase risk include cirrhosis, sickle cell anemia, multiple myeloma, chronic lung disease, and cancer. Organ transplantation recipients are also highly susceptible.

Table 1 Incidence of Pneumococcal Disease According to Age and Underlying Disease

Risk Factor

Cases of Pneumococcal Disease per 100,000 Population

    < 1 yr
    1 yr
    > 65 yr


Chronic medical conditions


Severe immunosuppression


The pneumococcus accounts for up to 40% of community-acquired pneumonias, causing or contributing to 40,000 deaths annually. The overall case-fatality rate of this pneumonia is 5% to 12%5; patients who are very young, elderly, alcoholic, leukopenic, infected with HIV, or infected with certain virulent pneumococcus serotypes (notably, type 3) are at the highest risk. Pneumococcal bacteremia develops in about 50,000 Americans annually, with a case-fatality rate of 20%—a rate that has not changed over the past 40 years.6 Pneumococcal otitis media is one of the most common childhood illnesses; the estimated incidence in the United States is seven million cases a year. Pneumococci cause about 3,000 cases of meningitis each year, with a case-fatality rate of about 21%.7

Epidemics of pneumococcal disease were common in the preantibiotic era, but sporadic cases are now the rule. Outbreaks have occurred, however, in shelters for the homeless, in jails, and in child care centers. Pneumococcal infection is typically community acquired, but it can occur nosocomially, particularly in chronically ill immunosuppressed patients; because of underlying diseases, the case-fatality rate of nosocomial pneumococcal bacteremia is 40%.8 The incidence of pneumococcal bacteremia is increased about 40-fold in patients with AIDS.


Pneumococci produce IgA protease, which cleaves the secretory immunoglobulin IgA1. IgA protease may be partly responsible for the high carrier rate and the respiratory tract pathogenicity of these bacteria. The organisms do not provoke an inflammatory response or cause clinical illness in the nasopharynx. Host defense mechanisms prevent penetration into more vulnerable areas, such as the paranasal sinuses, the middle ear, and pulmonary alveoli. Examples of these defenses include the cough, gag, and sneeze reflexes; the viscous mucus that lines the respiratory epithelium and traps bacteria; and ciliary action, which expels trapped particles. Pneumococcal infection often follows influenza9 or other viral infections of the upper respiratory tract, which impair host defenses by increasing the volume and decreasing the viscosity of secretions. Other factors that predispose to pneumococcal infection include dementia, seizure disorders, alcoholism, stupor, and other conditions that increase the likelihood that oropharyngeal contents will be aspirated into the lungs. Air pollution, cigarette smoking, heart failure, chronic obstructive pulmonary disease (COPD), and HIV infection increase the risk of pneumococcal infection, but diabetes does not.

Pneumococcal pneumonia begins with aspiration of small quantities of oropharyngeal secretions containing pneumococci. Normally, alveolar macrophages ingest and kill such bacteria, but in the absence of type-specific immunity, strains of pneumococci are protected from ingestion. The presence of pneumococci in the alveoli provokes a vigorous inflammatory response. The earliest manifestations are vasodilatation, increased vascular permeability, and exudation of edema fluid—the classic pathologic stage of congestion. Pneumococci can survive and even proliferate in the edema fluid. The organisms appear to float to adjacent alveoli, producing lobar consolidation. Within hours to days, polymorphonuclear leukocytes arrive and progressively pack the alveoli to produce the traditional stages of red and gray hepatization. However, only a minority of the pneumococci are phagocytosed and killed by the leukocytes in the first days of infection. Capsular polysaccharide antigen is often present in the serum and urine during these early stages.

After 5 to 7 days, type-specific anticapsular antibody appears. This results in a more efficient and enhanced opsonization that involves anticapsular antibody and the classical complement pathway. As pneumococci are ingested and killed by the polymorphonuclear leukocytes, the patient experiences a clinical crisis, marked by an abrupt fall in temperature and an increase in well-being; resolution commences. Because pneumococci only rarely produce significant tissue necrosis, healing is usually complete, and residual fibrosis is minimal.



Clinical presentation

Sir William Osler's classic description of pneumococcal pneumonia was recorded in 1892,10 only 10 years after the discovery of the pneumococcus, but it remains lucid and accurate today:

Abruptly, or preceded by a day or two of indisposition, the patient has a severe chill, lasting from ten to thirty minutes. In no acute disease is an initial chill so constant or so severe. The fever rises quickly. There is pain in the side, often of an agonizing character. A short, dry painful cough soon develops and the respirations are increased in frequency. When seen on the second or third day the patient presents an appearance which may be quite pathognomonic. He lies flat in bed, often on the affected side; the face is flushed, particularly the cheeks; the breathing is hurried; the alae nasi dilate with every inspiration; the eyes are bright, the expression is anxious, and there is a frequent short cough which makes the patient wince and hold his side. The expectoration is blood-tinged and extremely tenacious. The temperature rises rapidly to 104° or 105°. The pulse is full and bounding and the pulse-respiration ratio much disturbed. Examination of the lung shows the physical signs of consolidation—blowing breathing and fine rales. After persisting for from seven to ten days the crisis occurs, and with a fall in the temperature the patient passes from a condition of extreme distress and anxiety to one of comparative comfort.

Pneumonic sputum is viscid, tenacious, and blood-tinged. The gummous viscidity, together with the red blood-corpuscles in various stages of alteration, give pathognomonic characters to the sputa, unknown in any other disease. The rusty tinge becomes more marked as the disease progresses, and so tenacious is the expectoration that it has to be wiped from the lips of the patient, and a spitcup, half full, may be inverted without spilling. Microscopically, the sputum contains red blood-corpuscles in all stages of degeneration, alveolar epithelium, diplococci and other micro-organisms, cell-moulds of the alveoli, and, in some cases, small fibrinous casts of the bronchioles. The latter are sometimes plainly visible to the naked eye.

It should be emphasized, however, that not all patients with pneumococcal pneumonia present with a viral upper respiratory tract infection that is followed by the abrupt onset of fever, chills, pleurisy, dyspnea, and cough productive of purulent or blood-tinged sputum. If the patient has used antipyretics, fever can be modest or absent. The initial rigor may also be absent, but patients may occasionally have recurrent chills. In elderly or debilitated patients, confusion or stupor may be the presenting feature, far overshadowing pulmonary symptoms. In contrast, patients with chronic pulmonary disease may have rapidly progressive respiratory failure, which is disproportionately more severe than fever, sputum production, or other manifestations of infection. In rare cases, asplenic patients present with shock and hemorrhagic skin lesions that reflect disseminated intravascular coagulation. Because modest hyperbilirubinemia is common in pneumococcal pneumonia, right lower lobe pneumonia may masquerade as acute cholecystitis because of fever, jaundice, and right upper quadrant discomfort.

Laboratory tests

The key to the diagnosis of pneumococcal pneumonia is the Gram stain of a sputum smear, which typically reveals many polymorphonuclear leukocytes and abundant lancet-shaped gram-positive diplococci. If the patient cannot spontaneously produce an adequate sputum specimen, chest physiotherapy or nasotracheal suctioning should be considered. Sputum specimens should be promptly cultured on blood agar plates, preferably in 5% CO2 incubators. Because pneumococci are fastidious and fragile, the sputum culture may be negative in patients with clearly positive Gram stains. It is important to obtain blood cultures in all patients with suspected pneumococcal pneumonia. In 25% to 30% of these patients, the blood culture will be positive—a finding that confirms the diagnosis even if sputum cultures are negative. An immunochromatographic test can be used to detect pneumococcal polysaccharide in the urine of patients with pneumonia, but pneumococcal nasopharyngeal colonization can produce false positive results.11

Imaging tests

The classic physical and radiographic findings of lobar consolidation may be absent in patients with pneumococcal pneumonia. In fact, a bronchopneumonic pattern is radiographically more common than lobar consolidation. Dehydration may minimize pulmonary findings, and underlying chronic lung disease may predispose to patchy areas of pulmonary infiltration. Pleural effusions are relatively common and can occasionally obscure the underlying pulmonary parenchymal involvement.

Differential diagnosis

Other bacterial and nonbacterial pneumonias must be considered in the differential diagnosis of pneumococcal pneumonia [see 7:XX Pneumonia and Other Pulmonary Infections]. Less commonly, pulmonary edema, pulmonary emboli, atelectasis, or lung tumors are mistaken for pneumonia.

Complications and prognostic indicators

The complications associated with pneumococcal pneumonia have diminished markedly in the antibiotic era. Intrathoracic complications include pleurisy with sterile pleural effusion (common) and empyema (uncommon). Lung abscess is rare; if abscess or empyema occurs, it is likely to be caused by the heavily encapsulated type 3 pneumococcus or by a concomitant anaerobic infection. Purulent pericarditis is even rarer than abscess or empyema. Radiographic abnormalities often resolve slowly. About one third of patients have persistent consolidation at 1 month; although consolidation should resolve in all patients by 8 to 10 weeks, volume loss, pleural disease, and interstitial changes can persist for up to 4 months.

Bacteremia is by far the most common extrathoracic complication. Approximately 90% of patients with pneumococcal bacteremia have pneumonia, but patients occasionally present with pneumococcal bacteremias without an identifiable septic focus. Recurrent pneumococcal bacteremia may develop in patients with underlying diseases. Bacteremia is an adverse prognostic sign. The reported fatality rate is 11% to 36%, but it is probably higher in the elderly and in patients with severe underlying diseases. Metastatic infections such as meningitis, septic arthritis, peritonitis, and endocarditis are relatively uncommon; they appear to be more common and to occur at younger ages in African Americans than in whites.12

Several factors are associated with worse outcomes in pneumococcal pneumonia [see Table 2]. Interestingly, a lack of febrile response and a normal or low white blood cell count are readily measurable factors that are associated with worse outcome. Thus, although white blood cell counts of 25,000 to 30,000 with a left shift may be alarming, they indicate a favorable host response to infection.13

Table 2 Factors Associated with Adverse Outcomes in Pneumococcal Pneumonia

Infection with type 3 pneumococci
Multilobe infection
Extrapulmonary infection (meningitis, empyema)

Pneumococcal Infections in Persons Infected with HIV

The pneumococcus is the leading cause of invasive bacterial respiratory tract infections in HIV-positive persons. The clinical features, causative serotypes, antimicrobial resistance patterns, and mortalities of pneumococcal infection in HIV-positive patients are similar to those in HIV-negative patients. Severe infections, unusual extrapulmonary manifestations, and late relapses are more common in these patients, however. Pneumococcal vaccine should be administered as soon as possible after the diagnosis of HIV infection; despite the decreased efficacy of vaccine in patients with advanced AIDS, vaccination is a cost-effective preventive intervention at all stages of HIV infection.14

Other Pneumococcal Infections

Upper respiratory tract infections

Pneumococci spread from the nasopharynx to the upper respiratory tract to produce acute otitis media, especially in children. Acute mastoiditis, once a frequent sequela of acute otitis media, is now unusual. Pneumococci are a cause of acute purulent sinusitis in all age groups [see 7:XIX Bacterial Infections of the Upper Respiratory Tract].


Pneumococci can reach the central nervous system either by bacteremic spread from a pulmonary focus or by direct extension from otitis or sinusitis. Patients with skull fractures and cerebrospinal fluid rhinorrhea or otorrhea are particularly prone to recurrent attacks of pneumococcal meningitis [see 7:XXXVI Bacterial Infections of the Central Nervous System].

Septic arthritis and osteomyelitis

Pneumococci are a relatively common cause of acute septic arthritis, which results from bacteremic seeding; prosthetic joints may be involved [see 7:XV Septic Arthritis]. Pneumococcal osteomyelitis is uncommon.

Cardiac infections

Antibiotics have greatly reduced the incidence of pneumococcal pericarditis, which can result from direct extension of pneumonia or empyema or from hematogenous seeding. Pneumococcal endocarditis has also become uncommon in the antibiotic era.15

Postsplenectomy infections

Overwhelming postsplenectomy pneumococcal infection is an uncommon but important syndrome. In addition to patients who have undergone splenectomy, other patients at risk include those with sickle cell disease or other hemoglobinopathies that produce functional asplenia, as well as those with congenital asplenia. The syndrome is often marked by acute onset of fever, hemorrhagic skin lesions suggestive of disseminated intravascular coagulation or purpura fulminans, and shock. Hypoglycemia may be present. If therapy is not administered, death often occurs in less than 24 hours. Even when patients receive penicillin and cardiovascular support, mortality exceeds 50%. A similar syndrome can occur in healthy adults but is rare.16

Other infections

Primary pneumococcal peritonitis can occur in patients who have cirrhosis, nephrotic syndrome, systemic lupus erythematosus, or other host defects [see 7:XXI Peritonitis and Intra-Abdominal Abscesses]. In rare cases, pneumococci can infect the liver, the gallbladder, or pelvic organs; soft tissue infections and cellulitis caused by S. pneumoniae are also unusual.


For many years, all isolates of S. pneumoniae were penicillin sensitive; until 1965, most isolates in the United States were sensitive to less than 0.04 µg/ml penicillin. Subsequently, resistance to penicillin has become progressively more prevalent. Although pneumococci do not display plasmid-mediated penicillinase production, they can develop chromosomal mutations that confer resistance to penicillin by altering the affinity of the penicillin-binding proteins in their cell walls. Gradual remodeling of three or four of the penicillin-binding proteins in parallel produces a stepwise increase in the level of resistance. The DNA sequences responsible for resistance probably originated in other streptococcal species and were transferred to pneumococci by heterologous recombination.16

Pneumococci for which the minimum inhibitory concentration (MIC) of penicillin is less than 0.1 µg/ml are considered penicillin sensitive; those in which the MIC of penicillin is 0.1 to 1.0 µg/ml are considered intermediately resistant; and those for which the MIC is greater than 1.0 µg/ml are considered resistant. About 24% of pneumococcal organisms isolated in the United States are penicillin-nonsusceptible S. pneumoniae (PNSP), but in some areas, more than one third of S. pneumoniae isolates are resistant to penicillin.17 The prevalence of PNSP increases in patients who have been using antibiotics; it is highest in children younger than 6 years and in adults older than 65 years. Resistant strains can be spread from person to person, especially in closed population groups, such as those in day care centers, jails, and nursing homes. Infection with PNSP is associated with an increased likelihood of an adverse outcome.18

Pneumococci that are resistant to penicillin are often resistant to other antimicrobial drugs.17 First- and second-generation cephalosporins are generally ineffective against these organisms, but third-generation cephalosporins (particularly ceftriaxone and cefotaxime) and carbapenems are usually active. Erythromycin and the other macrolides are generally ineffective, as are clindamycin and trimethoprim-sulfamethoxazole (TMP-SMX); chloramphenicol has variable efficacy. Rifampin and vancomycin are active against virtually all isolates, but vancomycin tolerance has been identified and may become a concern in the future.19 Whereas many pneumococci have become resistant to the older fluoroquinolones, such as ciprofloxacin,20 newer agents in this class, such as levofloxacin, sparfloxacin, gatifloxacin, and moxifloxacin, are generally active against penicillin-resistant pneumococci. Levofloxacin resistance has been documented, however, especially in elderly nursing home patients, in patients with COPD, and in patients who had prior exposure to fluoroquinolones.20 Linezolid21is active against all pneumococci, as are ketolides, the still-investigational glycylcyclines, and daptomycin. Because of the increasing problem of PNSP, organisms isolated from clinical sources should be screened for penicillin resistance with a 1 µg oxacillin disk. Isolates with oxacillin inhibition zones of less than 19 mm should be studied further to determine the MICs of penicillin and other antimicrobial agents that are likely to be used in treatment.

Guidelines for treatment of patients with PNSP are being formulated.22 Pneumonia caused by PNSP strains can still be treated with penicillin G, although dosages must be in the range of 10 to 20 million units a day for average-size adults [see Table 3]. Ceftriaxone and cefotaxime are generally effective, but treatment failures in patients with meningitis have been reported. Until the results of susceptibility testing are available, it may be advisable to add vancomycin to the regimen of cefotaxime or ceftriaxone for patients with life-threatening pneumococcal infections such as meningitis. Vancomycin is an effective alternative for patients in whom treatment fails and for those who cannot tolerate cephalosporins; however, because of the limited CSF penetration of I.V. vancomycin in patients with pneumococcal meningitis, concurrent therapy with intrathecal vancomycin or I.V. rifampin may be advisable. The role of imipenem or meropenem in treating patients with PNSP is being explored. The newer fluoroquinolones (levofloxacin, sparfloxacin, gatifloxacin, and moxifloxacin) may be useful in adults, though none have indications in children. Linezolid has been approved for the treatment of pneumonia caused by PNSP, though no studies have been completed in patients with meningitis. To limit the spread of PNSP, physicians should avoid the inappropriate use of antibiotics and encourage the appropriate use of the pneumococcal vaccine.

Table 3 Antibiotic Treatment for Penicillin-Resistant S. pneumoniae




Penicillin G

10–20 million units/day

Despite resistance high-dose penicillin is effective for pneumonia


1 g I.V. q.d.
3 g I.V. q. 6 hr

Relative efficacy high, but cephalosporin resistance occurs*


1 g I.V. q. 6 hr

Variable efficacy; used sparingly because of toxicity


600 mg I.V. q. 24 hr

Used as part of combination therapy for meningitis only


1 g I.V. q. 12 hr

Intermediate efficacy; not justified for pneumonia but can be considered for meningitis*


500 mg I.V. q.d.
400 mg I.V. q.d.
400 mg I.V. q.d.

High efficacy for pneumonia; resistance to fluoroquinolones has been described


600 mg I.V. q. 12 hr

High efficacy for pneumonia; not studied for meningitis


800 mg p.o., q.d.

High efficacy; macrolide-resistant pneumococci are susceptible to telithromycin and other ketolides


1 g I.V. q. 8 hr

Could be used for cephalosporin-resistant pneumococci, but its effectiveness has not been studied systematically

*Triple-drug therapy with vancomycin, rifampin, and either ceftriaxone or cefotaxime may be necessary for pneumococcal meningitis if the minimum inhibitory concentration of either cephalosporin is > 4 mg/ml. Despite this resistance, vancomycin and cefotaximine demonstrate synergy.

Penicillin remains the drug of choice for susceptible pneumococci; cephalosporins are also active against such organisms but must be used with caution in patients allergic to penicillin. Other drugs that can be used to treat susceptible pneumococci include erythromycin and the new macrolides, ketolides, vancomycin, and the newer fluoroquinolones. Clindamycin has been a useful drug for treating a variety of pneumococcal infections, though inducible clindamycin resistance has been recently reported. TMP-SMX is a particularly good choice for patients with sinusitis or otitis; chloramphenicol may be useful for treating meningitis caused by susceptible pneumococci in patients who cannot tolerate penicillin, third-generation cephalosporins, or meropenem. Patients with pneumococcal pneumonia can be switched from intravenous to oral antibiotics once they are clinically stable, even if bacteremia was present.23


A vaccine containing 50 µg of purified capsular polysaccharide from 14 pneumococcal types was approved by the Food and Drug Administration in 1977, and an expanded vaccine containing 25 µg of polysaccharide from 23 pneumococcal types was released in 1983. These 23 types account for the great majority of serious pneumococcal infections in the United States, including 88% of invasive infections with penicillin-resistant strains.17 Because polysaccharide vaccines are not effective in children younger than 2 years, a protein-polysaccharide conjugate vaccine containing approximately 2 µg of polysaccharides from the seven most important strains conjugated to diphtheria toxoid was introduced for pediatric use in 2000.24

The 23-valent pneumococcal vaccination produces serum antibody titers that are protective in healthy adults; although most elderly people mount an adequate serologic response, some produce low antibody titers or functionally deficient antibodies. Patients with Hodgkin disease respond better to vaccination before staging laparotomy and splenectomy; chemotherapy markedly impairs the response to vaccination and accelerates the decline in antibody titers. Dialysis and transplant patients may respond to vaccination suboptimally,25 as do many patients with myeloma, lymphoma, leukemia, AIDS, and low serum vitamin B12 levels.

The pneumococcal vaccine is safe.26 Mild erythema and pain at the injection site may occur in up to one third of vaccine recipients, but fever, severe local reactions, and other serious side effects occur in fewer than 1%. Despite its immunogenicity and safety, the pneumococcal vaccine remains controversial because its efficacy has been difficult to demonstrate in certain population groups. In general, the evidence for efficacy has been stronger for the prevention of invasive pneumococcal disease than for nonbacteremic pneumonias27; the evidence is also stronger for the protection of low-risk persons than for the protection of elderly or debilitated patients.

Perhaps as a result of conflicting data, pneumococcal vaccine has not been well accepted, and only 35% of vaccine candidates have been immunized. Given the expense of large-scale field trials, the controversy is unlikely to be resolved in the near future, and recommendations for vaccine use must be formulated from the current information. Because of its safety and modest cost, the pneumococcal vaccine is cost-effective, even in the elderly.28 The vaccine is recommended for healthy adults older than 65 years and for patients with chronic cardiopulmonary disease, functional or anatomic asplenia (including sickle cell disease), Hodgkin disease, multiple myeloma, cirrhosis, alcoholism, renal failure, CSF leaks, immunosuppression, or HIV infection. Members of certain vulnerable population groups, such as Native Americans, are also candidates for the vaccine. When possible, the vaccine should be administered 2 weeks before elective splenectomy or chemotherapy [see CE:V Adult Preventive Health Care].

The pediatric conjugate vaccine is safe and effective29 and has reduced by 60% the rate of invasive infection caused by vaccine and vaccine-related serotypes in children younger than 5 years. Currently, it is recommended for all infants, as well as for children up to 60 months of age who are at increased risk for pneumococcal infection; vulnerable children older than 5 years should continue to receive the adult 23-valent vaccine.22

Although antibody titers wane,30 routine revaccination is not recommended. Apart from local reactions, however, revaccination after 5 or more years is safe31 and should be considered for adults who are at the highest risk for pneumococcal infection. The safety of the pneumococcal vaccine during pregnancy has not been evaluated; women who are at high risk for pneumococcal infection should be vaccinated before they become pregnant. The pneumococcal and influenza vaccines may be administered simultaneously at different sites when both are indicated.

Because of vaccine failures, additional strategies have been devised to protect vulnerable patients from pneumococcal infection. Long-term prophylaxis with orally administered low-dose penicillin protects children with sickle cell anemia from acquiring pneumococcal septicemia. It has also been recommended for young children with anatomic or functional asplenia. For asplenic adults, many physicians prefer vaccination with pneumococcal, meningococcal, and Haemophilus influenzae type b vaccines; these vaccines can be administered simultaneously. Vulnerable patients should be instructed to seek medical care at the first sign of infection and should be provided with ampicillin for self-administration if medical attention is not immediately available.

Additional work to enhance the efficacy of pneumococcal vaccines is needed. The development of a conjugated polysaccharide-protein vaccine for adults is particularly promising.


The streptococci, a large and diverse group of organisms widely distributed in nature, are part of the normal human flora. They vary in their pathogenic potential from many harmless species to a few very significant pathogens, such as S. pneumoniae and S. pyogenes.

Streptococci are gram-positive, round to ovoid cocci that can appear in pairs but more characteristically grow in chains of varying length. They are fastidious organisms requiring many nutrients. Although most are facultative anaerobes, some are obligate anaerobes. All species are nonmotile and nonsporulating and lack the enzyme catalase.

When grown on blood agar plates, most streptococci form small (1 to 2 mm), round, nonpigmented colonies. Streptococci are often characterized by changes in the appearance of the blood agar surrounding their colonies. Three types of reaction may occur: α-, β-, and γ-hemolysis. In α-hemolysis, green discoloration results from reduction of red blood cell hemoglobin and not from true hemolysis. In β-hemolysis, lysis of the red blood cells produces clearing of the blood agar. In the γ reaction, there is no change in the agar. Hemolysis alone, however, cannot be used for classification, because biologically dissimilar species can cause identical hemolytic reactions.

A far superior classification is the serologic system. In 1933, Rebecca Lancefield demonstrated that an antigenic carbohydrate could be extracted from the cell wall of streptococci. On the basis of chemical composition and immunologic reactivity of this carbohydrate, hemolytic streptococci can be divided into 18 groups (A through H and K through T). The species within each group tend to be biologically similar and to have a similar human pathogenic potential [see Table 4]. There are also separately classified nongroupable and anaerobic streptococci.

Table 4 Medically Important Streptococci and Enterococci


Representative Species



Human Habitat

Human Diseases


S. pyogenes

β (rarely γ)

Bacitracin sensitive; produces many extracellular enzymes and toxins


Many diseases (S. pyogenes is the major human pathogen)


S. agalactiae

β (α, γ)

May be bacitracin sensitive; hydrolyzes sodium hippurate

Genitourinary tract

Neonatal sepsis and meningitis
Puerperal infections
Urinary tract infections


S. equi

β (α, γ)

May be bacitracin sensitive; produces extracellular enzymes similar to those made by group A

Genitourinary tract
GI tract

Skin and wound infections


S. canis

β (α, γ)

Similar to group C

Similar to group C

Similar to group C


S. bovis
S. equinus

β, γ (α)

Heat resistant; bile resistant; NaCl sensitive; penicillin sensitive

GI tract
Genitourinary tract

Urinary tract infections


S. anginosus

β (α, γ)

Tiny colonies on blood agar; growth enhanced by 10% CO2

GI tract
Genitourinary tract

Brain abscess


S. sanguis

α (β, γ)

Bile resistant

GI tract


S. salivarius

γ (α)

Can grow at 45° C



Viridans streptococci, including S. mitis, S. mutans, and others


No group-specific carbohydrate antigen yet recognized



Anaerobic streptococci

Many, including the peptostreptococci

γ (β, α)

Obligate anaerobes; tiny colonies on blood agar; no group-specific carbohydrate

GI tract
Genitourinary tract

Lung abscess
Brain abscess
Soft tissue infections
Bone and joint infections


Enterococcus faecalis
E. faecium

β, γ (α)

Heat resistant; bile resistant; 6.5% NaCl resistant; penicillin resistant

GI tract
Genitourinary tract

Urinary tract infections
Wound infections


The group A streptococcus, S. pyogenes, ranks as one of the most important human pathogens. This organism can be recognized in the laboratory by the generous zone of β-hemolysis produced by most isolates. Up to 4% of group A streptococci, however, may be nonhemolytic while still retaining their full pathogenic potential. Because many other streptococci can produce β-hemolysis, a confirmatory test is necessary to identify β-hemolytic streptococci as group A. The traditional screening test is the bacitracin test; almost all group A streptococci are inhibited by very low concentrations of bacitracin. Definitive speciation of group A streptococci can now be accomplished rapidly through identification of the group A carbohydrate by immunologic techniques.

The group A streptococcus is structurally complex. The capsule consists of hyaluronic acid, which is similar to the hyaluronic acid of human connective tissue and is therefore nonantigenic. The cell wall is composed of group-specific carbohydrates, structural proteins, and mucopeptide. The group-specific carbohydrate is responsible for the Lancefield serologic grouping; in group A streptococci, this carbohydrate is a polymer of rhamnose and N-acetyl glucosamine. Three structural proteins are part of the cell wall. The most important is the M protein, which is a crucial marker of virulence because it impedes phagocytosis by polymorphonuclear leukocytes. Anti-M antibodies are protective because they opsonize the bacteria, thus promoting phagocytosis and killing. Among group A streptococci, there are more than 120 antigenically distinct M proteins and hence more than 120 M types. M typing of strains is useful for epidemiologic purposes. Clearly, a polyvalent M protein vaccine containing the most common M types would be necessary to provide protection in communities.

The third component of the group A cell wall is a mucopeptide composed of repeating units of N-acetyl glucosamine and N-acetyl muramic acid. This element can produce carditis in laboratory rabbits, but its role in human rheumatic fever is unknown. C5 peptidase is a cell-associated protein that cleaves the C5 component of complement, rendering it inactive. Streptolysin S is a cell-associated hemolysin that lyses red blood cells on contact. The final structural element is the lipid-protein protoplast membrane. This membrane has certain antigens that cross-react with antigens in human cardiac muscle. The importance of this phenomenon in the pathogenesis of rheumatic fever is uncertain.

Epidemiology of Group A Streptococcal Infections

The natural reservoirs of group A streptococci are the human pharynx and skin. Group A streptococci are transmitted by droplets, either from asymptomatic nasopharyngeal carriers or from persons with symptomatic pharyngitis.32 Both the carrier rate and the incidence of pharyngitis are highest in late winter and early spring, in temperate climates, and in school-aged children. The epidemiology of streptococcal pyoderma is quite different. It is most common in young children and has its peak in late summer and early fall. Group A streptococci can colonize normal skin and spread from person to person by direct contact. Nosocomial transmission can occur, resulting most often in postoperative wound infections or postpartum sepsis. On occasion, streptococcal infections can occur in epidemic fashion as a result of contaminated foods.

Several different M types of group A streptococci may be present in a community over the course of a year; thus, individuals (predominantly children from 5 to 10 years of age) may have three to four episodes of pharyngitis per year. In contrast, although 5% of adults may carry group A streptococci in their throats, symptomatic infection is uncommon, perhaps because of the acquistion of type-specific immunity from frequent pharyngeal infections earlier in life. The epidemiology of specific streptococcal infections varies widely, both geographically and temporally. Currently, scarlet fever and rheumatic fever are uncommon in the Western world, where they were very common as late as the 1940s. Interestingly, in northern climates, streptococcal pharyngitis has not decreased in prevalence over the past 100 years, even in developed countries. Around 1980, there was a dramatic increase in the incidence of severe invasive group A streptococcal infections such as necrotizing fasciitis, bacteremia, and the streptococcal toxic-shock syndrome (TSS); this higher incidence has persisted to the present time.33

Pathogenesis of Group A Streptococcal Infections

The ability of streptococci to adhere to mucosal and epidermal surfaces is important for the first step of pathogenesis, colonization. In the throat, colonization appears to be related to the elaboration of fibronectin-binding proteins (protein F). On the skin, group A streptococci adhere to CD44 on the surface of keratinocytes, largely through the bacteria's hyaluronic acid capsule. Adherence through these ligand pairs is sufficient to cause pharyngitis and impetigo, respectively. Although invasion of cell lines of mucosal and keratinocyte origin can occur in vitro, the role of cellular invasion in the pathogenesis of pharyngitis, impetigo, cellulitis, and erysipelas is unclear, because cellular and tissue destruction are either uncommon or are not sufficient to be clinically recognized. Still, an intracellular location could in part explain prolonged carriage and, in some cases, the failure of penicillin to eradicate infection. In necrotizing fasciitis, pneumonia, myositis, and other infections associated with bacteremia, direct invasion probably occurs. The organisms elaborate enzymes such as streptolysin O and S and nicotinamide adenine dinucleotide glycohydrolase (NADase), which can account for their characteristic ability to produce inflammation and tissue damage and to spread rapidly in tissues. Streptokinase catalyzes the conversion of plasminogen to plasmin and thus promotes fibrinolysis. Four distinct streptococcal deoxyribonuclease (DNase) enzymes have been identified; most streptococcal strains produce DNase B. Group A streptococci also produce hyaluronidase, which is known as spreading factor because of its ability to digest the hyaluronic acid of connective tissue. Except for streptolysin S, these enzymes are all antigenic. Although antibodies directed against these enzymes do not protect the patient against recurrent streptococcal infections, the tests for such antibodies can be used to obtain accurate serodiagnosis.

The second most common way in which group A streptococci cause disease is by elaboration of exotoxins, which are responsible for scarlet fever and streptococcal TSS. Three antigenically distinct pyrogenic exotoxins (formerly called erythrogenic toxins)—types A, B, and C—have been recognized for several decades, yet in the past 3 years, 12 additional pyrogenic exotoxins have been discovered. Elaboration of type A and type C pyrogenic exotoxin depends on lysogeny of the streptococci by a bacteriophage; on occasion, certain group C or group G streptococci or even staphylococci can produce a pyrogenic toxin. From the site of the streptococcal infection, which is usually in the pharynx, the toxin enters the circulation, producing the characteristic rash of scarlet fever through an unknown mechanism. In addition, these toxins function as superantigens that simultaneously attach to the major histocompatibility complex of antigen-presenting cells and to V beta regions of the T cell receptor, resulting in generation of inflammatory cytokines (tumor necrosis factor-α [TNF-α], interleukin-1 [IL-1], and IL-6), as well as the lymphokines TNF-β, gamma interferon, and IL-2. These cytokines play important roles in the pathogenesis of shock and organ failure. Antibodies to the toxins prevent development of a rash but do not protect against the underlying infection.

Finally, an immune response to antecedent infection with group A streptococci can produce nonsuppurative syndromes. Acute rheumatic fever (ARF) and acute glomerulonephritis begin 1 to 3 weeks after a group A streptococcal infection, usually when the infection is no longer active. Despite their common origins as poststreptococcal disorders, ARF and acute glomerulonephritis differ significantly in their epidemiologic and pathogenetic features.

Acute Group A Streptococcal Diseases

Upper respiratory tract infection

Group A streptococci are the most important cause of bacterial pharyngitis and tonsillitis.34 They also cause sinusitis, otitis, mastoiditis, cervical lymphadenitis, peritonsillar abscesses, and retropharyngeal abscesses. Rarely, streptococcal pharyngitis has been associated with streptococcal TSS [see 7:XIX Bacterial Infections of the Upper Respiratory Tract].


Streptococcal pneumonia is uncommon, accounting for less than 5% of all bacterial pneumonias. Although it is seen most frequently as a sequela of influenza, it can occur as a primary infection and may even produce epidemics in closed groups. Streptococcal pneumonia presents acutely as fever, chills, and productive cough. Its distinguishing features are pleuritic chest pain and rapidly progressive empyemas, both of which occur in approximately 60% of patients.

Lymphadenitis and lymphangitis

Group A streptococci are frequently responsible for suppurative lymphadenitis, particularly in patients with streptococcal infections of the respiratory tract or skin. In patients with a variety of soft tissue infections, red streaks that extend proximally are suggestive of group A streptococcal infection and are harbingers of lymphadenitis and subsequent bacteremia.

Wound infections

Group A streptococcal wound infections are characterized by early onset, often within 24 hours after surgery or injury. Patients present acutely with fever and systemic toxicity. Frequently, the infected wound appears relatively benign. A modest amount of thin serosanguineous fluid is often present in the wound. Gram stain of a smear of the fluid reveals gram-positive cocci in chains but a relative absence of acute inflammatory cells. If the infection is caused by strains of group A streptococci that produce pyrogenic exotoxins, a diffuse generalized erythema may develop (surgical scarlet fever).

Bone and joint infections

Group A streptococci are relatively common causes of infections of bones and joints, usually from bacteremic seeding.

Postpartum infections

The incidence of puerperal fever (childbed fever) diminished markedly from the 1850s, when it was common, to the 1990s, but it has increased in frequency over the past decade. The onset is abrupt, most often within 24 to 48 hours after delivery. Fever, chills, systemic toxicity, abdominal or pelvic pain, and a serosanguineous and odorless vaginal discharge are common; without treatment, endometritis may rapidly progress to pelvic peritonitis and bacteremia. These infections have also been associated with streptococcal TSS. Group A streptococci are the classic cause of puerperal fever, but other organisms, including group B streptococci and other streptococci and gram-negative bacilli, can cause infections with similar clinical features.


The incidence of group A streptococcal bacteremia appears to be rising. It often develops from a primary infection of skin, soft tissue, or a wound.35 The respiratory tract is the second most common source, but bacteremia after uncomplicated streptococcal pharyngitis is uncommon. In 18% of cases, no primary focus of infection can be identified. About 80% of patients have underlying problems, ranging from diabetes or malignancies to drug abuse. Osteomyelitis, septic arthritis, meningitis, and endocarditis36 are among the potential complications; overall mortality is 15%. Bacteremia is present in 60% of patients with necrotizing fasciitis and streptococcal TSS. Although meningitis can result from group A streptococcal bacteremia, this rare infection is more often a sequela of neurosurgical conditions.37

Skin and soft tissue infections

Group A streptococci are important causes of cutaneous infections, which can range from very superficial infections of the epidermis (e.g., impetigo) to deeper subcutaneous infections with systemic symptoms (e.g., erysipelas or cellulitis)38 to life-threatening processes (e.g., necrotizing fasciitis and gangrenous myositis) that require surgery and antibiotic therapy39 [see 2:VII Fungal, Bacterial, and Viral Infections of the Skin].

Necrotizing fasciitis

The widespread publicity that has been given to necrotizing fasciitis (previously known as streptococcal gangrene) has fueled popular concern about invasive group A streptococcal infections. An estimated 10,000 to 15,000 cases of invasive group A streptococcal infections occur each year in the United States; of these, necrotizing fasciitis occurs in 5% to 10%, with a case-fatality rate of about 30%. Most group A streptococci that cause invasive disease produce pyrogenic (formerly erythrogenic) toxin, but the genetic heterogenicity of causative strains does not support a clonal basis for the resurgence of invasive streptococcal infections.

Group A streptococcal necrotizing fasciitis is usually community acquired and sporadic in nature.39 It occurs in all age groups. Many patients have predisposing conditions, which may include chickenpox, trauma, diabetes, or alcoholism.40,41 In roughly 50% of patients, necrotizing fasciitis begins at the site of cutaneous penetrating trauma such as burns, insect bites, sliver injuries, abrasions, chickenpox vesicles, or surgical incisions. In the remaining cases, necrotizing fasciitis begins at the site of nonpenetrating deep trauma such as hematoma, ankle sprain, tendon rupture, or muscle tear. Most likely, group A streptococci translocate from the pharynx to the site of injury via the bloodstream. Because this type of infection begins deep within fascia and muscle, cutaneous changes are initially less prominent than pain, swelling, and systemic toxicity. Without aggressive therapy, however, the necrosis spreads to the skin and deep tissues. Computed tomography or magnetic resonance imaging can help define the depth and extent of infection, but in patients with preexisting trauma, infection may be difficult to distinguish from trauma alone. In addition to antibiotics and meticulous metabolic and circulatory support, prompt and aggressive surgical debridement is required; amputation may be necessary. Bacteremia and toxic streptococcal syndrome often complicate necrotizing fasciitis and are adverse prognostic features.

Diseases Caused by Toxin Production

Streptococcal toxic-shock syndrome

Since the first description of streptococcal TSS in 1987, the incidence of this disorder has remained relatively constant.42 Like staphylococcal TSS [see Staphylococci, below], streptococcal TSS appears to be mediated by a toxin, which in the case of the streptococcal form is probably a pyrogenic exotoxin. Like the staphylococcal toxin, the streptococcal toxin enters the circulation and functions as a superantigen, stimulating the release of host proteins that appear to mediate the shock syndrome.

The primary focus of infection is most often a soft tissue infection; respiratory infections are the next most common focus. Hypotension occurs in all patients and is often severe. Clinical features can include a generalized erythematous rash (10% of cases) that may undergo desquamation; acute respiratory distress syndrome (60% of cases); renal failure (80% of cases); and soft tissue necrosis, such as necrotizing fasciitis or myositis. Laboratory evidence of multiorgan involvement typically can be found and characteristically includes evidence of renal impairment, hepatic abnormalities, and disseminated intravascular coagulation, though clinical evidence of coagulopathy is rarely present.43

Management issues are complex and are reviewed in detail elsewhere,43 but aggressive fluid replacement is crucial; with it, hypotension resolves in 50% of patients. Patients with refractory hypotension may require replacement of albumin and correction of hypocalcemia. Brisk hemolysis can occur, particularly in the 60% of patients who have bacteremia, and may necessitate transfusion. Surgical consultation and aggressive surgical debridement of devitalized tissue are also crucial. Dialysis is necessary in patients who experience renal failure; ventilatory support is necessary in acute respiratory distress syndrome. Mortality in most streptococcal TSS series ranges from 30% to 70%. Intravenous immunoglobulin looked promising in initial case reports and in one observational study, but a subsequent double-blind study in Scandinavia did not demonstrate significant reduction in mortality or attenuation of necrotizing fasciitis, though the trial was cut short because of low enrollment.44 Earlier diagnosis, aggressive fluid resuscitation, and appropriate antibiotics may be responsible for the somewhat lower mortalities reported in more recent series.43

Scarlet fever

The incidence of scarlet fever has declined sharply in the antibiotic era. The initial symptoms are fever and sore throat. Within 1 to 5 days, the characteristic fine, red, sandpaperlike eruption appears, often beginning on the chest and rapidly spreading to other parts of the body. Although the tongue and buccal mucosa are classically involved, the perioral area may be spared, thus accounting for the typical circumoral pallor. The rash is caused by hyperemia and capillary damage produced by pyrogenic toxin. In areas of trauma, such as the antecubital fossae, punctate hemorrhages (Pastia sign) may occur. Nausea and vomiting may be present, and fever and prostration may be severe. Desquamation of skin and mucous membranes is prominent during healing; one characteristic feature is the strawberry tongue. Therapy is the same as that for the underlying streptococcal infection.

Delayed Diseases Caused by the Host Immune Response

Group A streptococci can produce disease by a mechanism that is rarely a feature of other infections—a host immune response that subsequently produces tissue damage. ARF and acute glomerulonephritis are the major syndromes produced in this fashion; both are characterized by a latent period between the streptococcal infection and its inflammatory consequences. Erythema nodosum is a third poststreptococcal syndrome, but it is less specific and less serious than ARF and acute glomerulo nephritis.

Acute rheumatic fever

ARF is strictly a sequela of streptococcal pharyngitis. Although many strains of group A streptococci can provoke ARF, the syndrome most often follows pharyngeal infection by certain M protein types (especially M5, M18, and M3), which are heavily encapsulated, mucoid organisms. In about half the cases, the antecedent pharyngitis is clinically silent, but all patients with ARF have serologic evidence of a recent group A streptococcal infection. The risk of ARF after untreated streptococcal pharyngitis is less than 3%; the latent period between the pharyngitis and the onset of ARF averages 18 days.

Although common in developing nations, ARF is now quite uncommon in industrialized nations. However, the occurrence of several outbreaks in the United States in the 1980s demonstrates that ARF can still be an important problem in affluent societies. The disease occurs most often in children from 5 to 15 years of age; adults may present with atypical features, including synovitis without carditis.

Despite intensive study, the pathogenesis of ARF is unclear. The most widely held theory proposes that a genetically susceptible host develops an autoimmune response to epitopes in the organism that are cross-reactive with epitopes in tissues of the heart, joints, skin, or CNS.45 Alternatively, streptococcal toxins or immune complexes may create alterations in tissue antigens that in turn provoke an autoimmune response that damages host tissues.

There are five major clinical manifestations of ARF: carditis, polyarthritis, chorea, subcutaneous nodules, and erythema marginatum. Carditis develops in about 60% of patients; it involves the endocardium, myocardium, and pericardium. The typical manifestations of rheumatic carditis are sinus tachycardia (sometimes with first-degree heart block), mitral regurgitation, a pericardial friction rub, and cardiomegaly; congestive heart failure indicates severe carditis. Although most cases of carditis resolve within 3 months, patients with moderate to severe carditis or recurrent ARF are at risk for the late manifestation of mitral valve or aortic valve scarring [see 1:XI Valvular Heart Disease]. Polyarthritis develops in about 70% of patients with ARF. It is characteristically a migratory arthritis involving the large joints of the extremities; it resolves without sequelae in days to weeks.46 On rare occasions, adults may develop a persistent arthropathy of the hands and feet. Chorea, subcutaneous nodules, and erythema marginatum are all self-limited; each occurs in fewer than 10% of children with ARF and only very rarely in adults.

The minor manifestations of ARF are fever, arthralgias, and inflammation. The inflammation in such cases is evidenced by elevated erythrocyte sedimentation rates and C-reactive protein levels.

The diagnosis of ARF is made on the basis of clinical features. The classic Jones criteria include the presence of either two major manifestations or one major and two minor manifestations, as well as laboratory evidence of a recent streptococcal infection (e.g., a positive throat culture or rising antistreptococcal antibody levels) [see Table 5].47

Table 5 Revised Jones Criteria for the Diagnosis of Acute Rheumatic Fever47

Major manifestations

Erythema marginatum
Subcutaneous nodules

Minor manifestations

  Previous rheumatic fever or rheumatic
      heart disease
  Elevated acute-phase reactants
    Erythrocyte sedimentation rate
    C-reactive protein level
  Prolonged PR interval

Supporting evidence of
  streptococcal infection

Increased titer of antistreptococcal
    antibodies (e.g., anti-streptolysin O)
Positive throat culture for group A

Note: The presence of two major manifestations or of one major and two minor manifestations indicates a high probability of acute rheumatic fever, if there is supporting evidence of group A streptococcal infection.

Treatment of ARF focuses on eradication of streptococcal pharyngitis and reduction of inflammation [see Table 6]. Most clinicians recommend a course of penicillin or, as an alternative, other antistreptococcal antibiotics, even if the throat culture is negative at the time ARF is diagnosed. Anti-inflammatory therapy includes aspirin and bed rest until inflammatory symptoms resolve; corticosteroids may have a role for patients with severe carditis.

Table 6 Drug Treatment for Acute Rheumatic Fever93





Acute treatment:
  Benzathine penicillin G, 1.2 million units I.M.
  Penicillin V,*500 mg p.o., b.i.d.-t.i.d.
  Penicillin VK, 250 mg p.o., b.i.d.
  Benzathine penicillin G, 1.2 million units I.M. q. 4 wk

Acute treatment is given for patients with acute rheumatic fever who have a positive throat culture for group A Streptococcus; benzathine penicillin can be given if compliance is uncertain; prophylaxis for patients with history of acute rheumatic fever is begun immediately after the acute episode and continued until the patient reaches young adulthood


1 g p.o., q.d

For prophylaxis in patients allergic to penicillin


Acute treatment: erythromycin ethylsuccinate,
  250 mg q.i.d.
Prophylaxis: 250 mg p.o., b.i.d.

Acute treatment or prophylaxis for patients allergic to penicillin


90–100 mg/kg/day for 2 wk, then 60–70 mg/kg/day
  for 6 wk

Anti-inflammatory agents do not alter the development of chronic heart disease but reduce symptoms; blood salicylate levels of 20 mg/100 ml are desirable


40–60 mg q.d. for 2 wk, then taper off over 3 wk

Prednisone or a nonsteroidal anti-inflammatory drug can be used when symptoms are not controlled with aspirin alone

*Phenoxymethyl penicillin.

ARF can be prevented by adequate treatment of streptococcal pharyngitis, even if antibiotics are delayed for up to 9 days after the onset of pharyngitis. Patients with a history of ARF are particularly vulnerable to recurrent attacks. Consequently, they should receive continuous prophylaxis for at least 5 years with daily oral penicillin (250 mg twice a day) or monthly injections of 1.2 million units of benzathine penicillin G. Oral sulfadiazine or erythromycin may be administered to patients allergic to penicillin. Prophylaxis can be discontinued when young patients who are at low risk for recurrence reach adulthood or when small children are no longer in the household.

Acute glomerulonephritis

Acute poststreptococcal glomerulonephritis appears to result from the deposition of circulating antigen-antibody complexes and complement in renal glomeruli [see 10:V Glomerular Diseases]. Only group A streptococci of certain M protein types can produce glomerulonephritis; it is unknown why only approximately 15 of the more than 80 M types are nephritogenic. Acute glomerulonephritis can follow either streptococcal pharyngitis or pyoderma but is more common after skin infections.

The incidence of acute glomerulonephritis after infection with nephritogenic streptococci varies but may reach 10% to 15% in certain epidemic situations. The prognosis is generally good, especially in children, and recurrences are uncommon; therefore, penicillin prophylaxis is not indicated. In fact, it is not certain that penicillin therapy for streptococcal pharyngitis or pyoderma affects the incidence of subsequent acute glomerulonephritis.

Erythema nodosum

The painful pretibial nodules of erythema nodosum can develop after streptococcal infections of the skin, pharynx, or other sites. However, this nonsuppurative sequela may be associated with other infectious processes, such as tuberculosis or systemic mycoses, and with noninfectious processes, such as inflammatory bowel disease or hypersensitivity reactions.

Diagnosis of Group A Streptococcal Infections

The diagnosis of group A streptococcal infections depends on recognizing the clinical syndromes and culturing the bacterium from appropriate specimens. Streptococcal pharyngitis can be diagnosed rapidly by detecting streptococcal antigens on throat swab specimens, but results are dependent on the type of rapid test and the expertise of the laboratory technician [see 7:XIX Bacterial Infections of the Upper Respiratory Tract]. Rapid tests have incomplete sensitivity, so if a rapid test is negative, a throat culture should be obtained. Serologic tests may be extremely useful, albeit retrospectively, in documenting recent streptococcal infection. Many of the enzymes and toxins produced by group A streptococci are antigenic, and a variety of antibody tests are available. The most widely used is the anti-streptolysin O titer, which is elevated after most respiratory tract infections; anti-DNase B titers are also elevated after most group A streptococcal infections [see Table 7].

Table 7 Laboratory Tests for Streptococcal Pharyngitis94


Sensitivity (%)

Specificity (%)


Rapid tests



Results are operator dependent; only qualified personnel should perform the test

Anti-streptolysin O titer



Elevation may also occur after infection with group C or G streptococci; a single titer > 200 Todd units/ml is significant; a rise in titer of > 2 dilutions between acute and convalescent sera is significant; responses are good after throat infections and poor after skin infections

Anti-DNase B titers



Consistent titer rises after throat or skin infection

Anti-hyaluronidase titers



Consistent titer rises after throat or skin infection


Treatment of Group A Streptococcal Infections

Despite the widespread use of penicillin, group A streptococci continue to be uniformly sensitive to this agent, which remains the drug of choice. The dose schedule and duration vary enormously. The standard therapy for pharyngitis is 10 days of low-dose oral penicillin (or a single I.M. dose of benzathine penicillin G), but a large trial suggested that 5 days of therapy with various β-lactams or macrolides may be as effective.48 Similarly, although oral penicillin is usually administered four times a day, a meta-analysis found that twice-daily dosing is as effective for children.49 In contrast, more invasive streptococcal infections require much more intensive treatment, such as prolonged high-dose I.V. therapy for osteomyelitis or endocarditis. In patients who are allergic to penicillins, cephalosporins are effective but must be used with caution because of the risk of allergic cross-reactivity. Erythromycin, vancomycin, and clindamycin are generally excellent alternatives, although occasionally, clinical isolates are resistant to these agents. Erythromycin-resistant strains have become prevalent, especially in Japan and Finland and, more recently, in the eastern United States. Reducing the use of macrolide antibiotics can control this problem. Many strains of group A streptococci are now resistant to tetracyclines and TMP-SMX. The newer fluoroquinolones levofloxacin, sparfloxacin, gatifloxacin, and moxifloxacin are active against group A streptococci, but these agents are not indicated in children. In patients with necrotizing fasciitis and streptococcal TSS, clindamycin's ability to suppress streptococcal toxins may make it preferable to penicillin.


Streptococci belonging to all the Lancefield groups can cause disease ranging in severity from urinary tract infections to meningitis, bacteremia, and endocarditis [see Table 4]. None of the non-group A organisms, however, have been implicated in rheumatic fever.

Groups C and G Streptococci

Groups C and G streptococci have many similarities to group A streptococci, although they are much less commonly implicated as pathogens. Most strains of groups C and G are β-hemolytic, and some are bacitracin sensitive. Bacteria in both groups can produce many of the same enzymes and toxins produced by group A streptococci, including streptolysin O, streptokinase, NADase, DNase, hyaluronidase, and pyrogenic toxin. The group-specific carbohydrate of group C is similar to that of group A, and groups A, C, and G all have cross-reacting cell membrane antigens.

First recognized as animal pathogens, groups C and G streptococci may be part of the resident flora of the pharynx, skin, genitourinary tract, and gastrointestinal tract.50 Group C streptococci can cause pharyngitis in adults. Groups C and G streptococci can cause skin and wound infections, puerperal sepsis, bacteremia, endocarditis, pericarditis, septic arthritis, meningitis, scarlatiniform eruptions, myositis, and TSS.51 In rare instances, groups C and G streptococci cause poststreptococcal glomerulonephritis. Groups C and G streptococci are sensitive to penicillin, which is the drug of choice, and to erythromycin, vancomycin, and other antibiotics.

Group B Streptococci

Group B streptococci are important human pathogens.52 Most group B streptococci are β-hemolytic; up to 20% are sensitive to low-dose bacitracin. They can be presumptively separated from group A streptococci on the basis of their ability to hydrolyze sodium hippurate, but definitive identification depends on immunologic identification of the group B carbohydrate.

Group B streptococcal infections are the leading cause of bacterial disease and death in newborns; maternal infections can also occur in the peripartum period.52 The organisms reside in the vagina and infect neonates during passage through the birth canal; premature labor, prolonged membrane rupture, and intrapartum fever are risk factors for neonatal infection. The intrapartum administration of antibiotics to women at high risk can prevent neonatal infection; screening during the third trimester of pregnancy and treatment with penicillin have substantially reduced the incidence of perinatal and peripartum group B streptococcal infections.53

Group B streptococci can cause a broad range of infections in nonpregnant adults. Persons of any age can be affected, but elderly patients are most vulnerable.54 Other risk factors include genitourinary disorders, diabetes, cancer, HIV infection, and peripheral vascular disease. In addition to bacteremia, infections caused by group B streptococci include urinary tract infections, septic arthritis, endocarditis, meningitis, pulmonary infections, and infections of the skin, soft tissue, and wounds. Necrotizing fasciitis and streptococcal TSS55 have also been reported.

Although group B streptococci are sensitive to clinically achievable levels of penicillin, the MICs are somewhat higher than those for group A streptococci. The great majority of group B streptococci are sensitive to ampicillin, cephalosporins, vancomycin, erythromycin, and clindamycin but are resistant to tetracycline.

Group D Streptococci

Group D streptococci (S. bovis and S. equinus) can be identified in the laboratory by their ability to resist heat and bile but not 6.5% NaCl broth. Both species reside in the human GI and genitourinary tracts. S. equinus rarely causes human disease, but S. bovis is a relatively common cause of endocarditis, which may be more severe than the subacute bacterial endocarditis caused by viridans streptococci.56 S. bovis bacteremia is significantly linked to carcinoma of the colon; therefore, the GI tract of any patient with this infection should be aggressively evaluated. S. bovis is sensitive to penicillin and other antibiotics and responds well to therapy with penicillin alone.

Nongroupable Streptococci

Up to 30% of streptococcal isolates are not groupable under the currently available Lancefield antisera classification. These organisms are capable of producing a great variety of pyogenic infections, among which subacute bacterial endocarditis stands out as the most important. Nongroupable streptococci can also cause bacteremia and pneumonia, especially in immunosuppressed patients57; meningitis is rare.58Members of the S. milleri group have a propensity to produce abscesses, particularly in the brain or liver.59 Classically, subacute endocarditis is caused by viridans streptococci—nongroupable, α-hemolytic bacteria that are part of the normal flora of the oropharynx [see7:XVIII Infective Endocarditis]. Many species of viridans streptococci can cause endocarditis; S. sanguis, S. mutans, S. mitis, and members of the S. milleri group are commonly responsible. Infections caused by members of the S. milleri group may be more destructive than those caused by the other species. About 10% of viridans streptococci are nutritionally variant, requiring supplementary pyridoxal to grow on agar plates. Up to one third of these organisms require serum penicillin levels of greater than 0.1 µg/ml to inhibit growth. Although this level is higher than the levels that kill most other streptococci, this concentration can be easily achieved clinically. Synergistic penicillin-aminoglycoside therapy has been recommended for endocarditis caused by these organisms. Penicillin-tolerant and penicillin-resistant strains of viridans streptococci have been increasingly recognized; vancomycin or penicillin-aminoglycoside therapy has been recommended for endocarditis caused by penicillin-resistant streptococci.

Anaerobic Streptococci

Various anaerobic and microaerophilic streptococci reside in the oropharynx, the GI tract, and the genitourinary tract. These organisms cause a great variety of infections, including aspiration pneumonia, lung abscess, empyema, sinusitis, brain abscess, bone and joint infections, and skin and wound infections. Anaerobic streptococci often participate in these processes, along with other anaerobic or aerobic bacteria. Foul-smelling pus is characteristic of these infections, and gas may be present in soft tissues. The anaerobic streptococci are penicillin sensitive.


Although they were long considered to be streptococci, the enterococci have been reclassified into their own genus, Enterococcus. Of the infections caused by enterococci, 80% to 85% are caused by E. faecalis; most of the others are caused by E. faecium, but E. durans, E. avium, and other species can also cause disease in humans. Morphologically indistinguishable from streptococci and immunologically similar to members of group D streptococci, the enterococci are metabolically unique in their ability to resist heat, bile, and 6.5% NaCl broth. Unlike streptococci, enterococci are uniformly penicillin resistant. Both major enterococcal species reside in the GI and genitourinary tracts, and both can cause urinary, abdominal, and disseminated infections, including endocarditis on preexisting valvular lesions.


The primary reason enterococci have emerged as major human pathogens is that these organisms are resistant to many antibiotics. Enterococci are intrinsically resistant to penicillin because they have unique penicillin-binding proteins that permit cell wall synthesis to proceed even in the presence of β-lactam antibiotics. In addition, enterococci that produce β-lactamase have been recognized; although these strains are still relatively uncommon, they may pose problems in the future. Enterococci can also demonstrate low-level resistance to aminoglycosides, but synergistic therapy with a penicillin and an aminoglycoside has been successful against most species. However, high-level resistance to the aminoglycosides streptomycin and gentamicin has emerged during the past 20 years. More recently, vancomycin-resistant enterococci (VRE) have been recognized as nosocomial pathogens.60 VRE colonization of the GI tract and skin often precedes overt infections. Colonization and infection are usually hospital-acquired; they are most common in patients who have prolonged hospitalizations or serious underlying diseases or who have undergone previous therapy with multiple antibiotics. Antibiotics that are active against anaerobes are particularly likely to promote intestinal colonization.61 VRE are often carried by medical personnel and can spread from patient to patient. Because they often occur in very ill patients and are so difficult to treat, VRE infections pose a major threat. The mortality associated with VRE bacteremia approaches 45%.59


Enterococci, alone or with other enteric organisms, are relatively common causes of urinary tract infections, wound infections, and peritonitis and intra-abdominal abscesses. Enterococci have become an increasingly prominent cause of bacteremia,62 which usually originates from a focus in the urinary tract or abdomen; the incidence of nosocomial bacteremias caused by these organisms is also increasing, particularly in patients who have received cephalosporins or other broad-spectrum antibiotics. Enterococcal endocarditis may affect normal, diseased, or prosthetic valves and may pursue an acute destructive course. Enterococcal meningitis is much less common but may be very difficult to treat. Enterococcal infections are most common in persons with underlying genitourinary or GI disease, in the elderly, and in debilitated persons. Enterococci have become an important cause of disease in hospitalized patients; they are now the second most common nosocomial pathogen in the United States, occurring less commonly than Escherichia coli but more commonly thanPseudomonas aeruginosa and Staphylococcus aureus.


Changing resistance patterns will necessitate changes in antibiotic therapy for patients with enterococcal infections [see Table 8]. Traditionally, ampicillin has been the drug of choice for en terococcal urinary tract infections, largely because of the high concentrations of drug excreted in the urine. In penicillin-allergic patients, nitrofurantoin, a fluoroquinolone, or TMP-SMX may be effective for the treatment of simple urinary tract infections; however, because in vitro susceptibility testing of enterococci is often misleading and because resistance to these agents may rapidly emerge, patients should be continually evaluated through assessment of their clinical response and by use of follow-up cultures.

Table 8 Antibiotic Treatment for Enterococcal Infections95




Ampicillin + gentamicin

Ampicillin: 2 g I.V. q. 4 hr
Gentamicin: 1 mg/kg I.V. q. 8 hr

Indicated in patients with endocarditis or meningitis*

Vancomycin + gentamicin

Vancomycin: 1 g I.V. q. 12 hr
Gentamicin: 1 mg/kg I.V. q. 8 hr

For penicillin-allergic patients*


2 g I.V. q. 4 hr

For infections caused by enterococci that produce
  β-lactamase; these appear to be rare


1 g I.V. q. 12 hr

For infection caused by strains resistant to aminoglycosides


600 mg q. 12 hr I.V. or p.o.

For infection caused by vancomycin-resistant strains


7.5 mg/kg I.V. q. 8 hr

For infection caused by vancomycin-resistant strains; not effective against E. faecalis


100–200 mg I.V. or p.o., q. 12 hr

For infection caused by vancomycin-resistant strains; experience very limited


500 mg I.V. or p.o., q. 6 hr

For infection caused by strains resistant to vancomycin, aminoglycosides, and penicillin


100 mg p.o., q. 12 hr

For urinary tract infections caused by vancomycin-resistant strains

*If high-level resistance to gentamicin (500–2,000 µl) is detected, perform susceptibility testing to streptomycin; substitute streptomycin if sensitive.

Deep tissue and bloodstream infections have traditionally been treated with ampicillin and gentamicin, with vancomycin, or with vancomycin and gentamicin. Although these agents are still appropriate for most enterococcal infections, all clinically significant isolates should be subjected to testing for β-lactamase production, high-level aminoglycoside resistance, and vancomycin resistance to determine whether an alternative therapy is necessary. Infections caused by enterococci that produce β-lactamase can be treated with an antimicrobial agent that combines a penicillin with a β-lactamase inhibitor. Infections caused by strains that are highly resistant to aminoglycosides can be treated with vancomycin. Infections caused by vancomycin-resistant strains are the most difficult to treat; linezolid has been particularly helpful, and quinupristin-dalfopristin is also effective.63 Doxycycline, chloramphenicol, and, for urinary tract infections, nitrofurantoin have also been used. In all cases, removal of infected indwelling catheters and foreign bodies and drainage of septic foci are essential. It is also essential to limit the spread of VRE by intensive surveillance, infection-control measures,64 and prudent use of vancomycin and other antibiotics.


Staphylococci are nonsporulating, nonmotile gram-positive cocci that have an average diameter of 1 µm. Microscopically, staphylococci tend to be larger and rounder than streptococci. Because cell division occurs on three planes, these organisms are typically found in grapelike clusters and tetrads, as well as in pairs and sometimes in short chains. When grown on blood agar, staphylococci form small (1 to 2 mm), smooth, round colonies that are often pigmented and may be surrounded by a zone of β-hemolysis.

Staphylococci are very hardy organisms and can withstand much more physical and chemical stress than pneumococci and streptococci. For example, staphylococci resist drying, withstand 10% NaCl broth, and will survive and even replicate at temperatures between 10° and 45° C. Because staphylococci are facultative anaerobes, they will grow in the presence or absence of oxygen. Staphylococci are catalase positive.

Of the species of staphylococci, S. aureus is by far the most important human pathogen. S. aureus can be tentatively identified in the laboratory on the basis of its production of a golden-yellow pigment, which may not be apparent until after 24 hours of growth. Because some strains are nonpigmented, definitive identification of S. aureus depends on its ability to produce the enzyme coagulase and to ferment mannitol. The most important coagulase-negative organism is S. epidermidis, which is universally present as part of the normal skin flora but can produce disease in certain circumstances, such as infection of an indwelling prosthesis. Another important coagulase-negative species is S. saprophyticus, which causes urinary tract infections.


The cellular structure of S. aureus is complex.65 Most strains have polysaccharide microcapsules.

The cell wall of S. aureus is structurally similar to that of group A streptococci: both have a carbohydrate antigen, a protein component, and a mucopeptide. The carbohydrate antigen is a teichoic acid, which in S. aureus is a polymer of N-acetylglucosamine and polyribitol phosphate. Antibodies to teichoic acid can be detected in normal human serum, and elevated antibody titers are present in patients with deep-seated staphylococcal infections. Teichoic acid has no established role in virulence, and antibodies to this carbohydrate are not protective. The protein component of the cell wall includes protein A, which reacts with IgG of normal human serum. Protein A interacts with the Fc component rather than the Fab component of IgG and hence is not a true antigen. Protein A may be antiphagocytic, but its role in virulence has not been clearly established. The cell wall mucopeptide of staphylococci is structurally similar to the mucopeptide of other gram-positive bacteria.

Epidemiology of Staphylococcal Infections

  1. aureusis an extremely widespread organism; it can be cultured from environmental surfaces, clothing, bedding, and the air, although transmission of staphylococci by airborne droplets or fomites does not appear to be a major route of infection. Humans—either asymptomatic carriers or persons with staphylococcal lesions—are the reservoir of infection. Nasal carriage is common; rates of nasal carriage range from 15% to more than 50% and are increased in hospitalized patients, dialysis patients, intravenous drug users, and insulin-injecting diabetic patients. Other relatively common sites of colonization include the rectum, the perineum, and the pharynx. S. aureuscan often be cultured from normal skin, but skin colonization is probably brief, with repeated repopulation occurring from the nose. It is unclear why some persons are prolonged carriers, others are transient carriers, and others resist colonization and why some persons carry many organisms, whereas others have low colony counts. Person-to-person transmission is more likely from nasal carriers of large numbers of organisms and from persons with active staphylococcal infections of the skin or respiratory tract. Nasal carriage increases the risk of bacteremia and of skin and wound infections, including recurrent furunculosis66; upper respiratory tract viral infections appear to increase bacterial shedding by S. aureus nasal carriers. Periodic application of mupirocin ointment in nasal carriers reduces the rate of carriage and lowers the risk of skin infections.

Minor staphylococcal infections are extremely common. It is the rare person who has not at some time experienced a staphylococcal furuncle (boil), paronychia, or hordeolum (sty). Serious staphylococcal infections generally require a predisposing insult to host defenses. Most often, this takes the form of skin disease, trauma, or a viral infection of the respiratory tract, especially influenza or measles. Other predisposing factors include foreign bodies, liver disease, neoplasia, diabetes, renal failure, defects in leukocyte or immunoglobulin function, narcotics addiction, and broad-spectrum antibiotic therapy.

Unlike the incidence of infections from other gram-positive cocci, the incidence of serious staphylococcal infection increased sharply after the introduction of antibiotics. Much of this increase can be attributed to the development of antibiotic resistance. When penicillin was first introduced, fewer than 10% of staphylococci were penicillin resistant; this percentage has increased steadily and now includes the great majority of staphylococci. During the late 1950s, staphylococcal infection reached epidemic proportions, and hospital-acquired infections were a particularly grave problem. With the introduction of methicillin in 1959 and other penicillinase-resistant antibiotics in the 1960s, the incidence of nosocomial staphylococcal infection temporarily declined. Unfortunately, however, methicillin-resistant staphylococci have slowly but progressively emerged as major nosocomial pathogens in many parts of the world. In addition, in recent years, community-acquired methicillin-resistant S. aureus (MRSA) infections have increased in prevalence in many regions of the United States, Japan, and Southeast Asia. Vancomycin-intermediate and vancomycin-resistant strains have appeared in both Japan and the United States.67

Pathogenesis of Staphylococcal Infections

The earliest tissue response in staphylococcal infection is acute inflammation with a vigorous exudation of polymorphonuclear leukocytes. Vascular thrombosis and tissue necrosis quickly lead to abscess formation. As a result of the development of a fibrin meshwork and, later, fibroblast proliferation, these abscesses become walled-off zones of loculated infection and tissue destruction, with dying leukocytes and viable bacteria at the center. Fibrosis and scarring are often prominent in healing.

Most strains of S. aureus produce a variety of extracellular products, including both enzymes and toxins, that may account for the tendency to produce burrowing, destructive, localized infections. The enzyme coagulase causes plasma to clot, thus promoting the fibrin meshwork that contributes to abscess formation. Staphylococci can also produce lipase, protease, hyaluronidase, and DNase, which can add to tissue damage. Another important enzyme is penicillinase. Because penicillinase has no role in pathogenicity, staphylococci that produce penicillinase are no more virulent than non-penicillinase-producing strains. Nevertheless, this enzyme is clinically and epidemiologically important because it hydrolyzes the β-lactam ring of penicillin, thereby inactivating the molecule. The production of penicillinase is controlled by plasmids, or episomes, which are extrachromosomal DNA molecules that replicate during cell division. Unlike the R factors of gram-negative bacilli, however, the plasmids responsible for penicillinase production do not usually mediate resistance to multiple antibiotics.

Of even greater interest are the nonenzymatic toxins produced by S. aureus. α-Toxin is a cytotoxin that produces pores in cell membranes, thereby altering their permeability and resulting in cell damage or death. α-Toxin damages red and white blood cells and activates platelets. Injection of α-toxin into animals can produce dermal necrosis and contraction of vascular smooth muscle, leading to tissue ischemia. Another potential virulence factor is leukocidin, which consists of two leukotoxic proteins that are capable of disrupting lysosomal membranes. Occasionally, strains of staphylococci produce exfoliatin, which causes the epidermolysis characteristic of staphylococcal scalded skin syndrome (SSSS). Some strains of staphylococci can also produce one of four antigenically distinct enterotoxins that cause the vomiting and diarrhea characteristic of staphylococcal food poisoning. In rare instances, staphylococci produce an erythrogenic toxin that causes scarlet fever. Finally, a staphylococcal exotoxin, toxic-shock syndrome toxin-1 (TSST-1), appears to be responsible for menstrually related staphylococcal TSS. Interestingly, staphylococcal TSS associated with nasal packing and other surgical wound infections is mediated by one of the enterotoxins, most commonly enterotoxin-B.

Host resistance and immunity to staphylococci are poorly understood. The importance of the granulocyte is supported by the susceptibility to staphylococcal infections seen in patients with neutropenia or various disorders of neutrophil function, such as chronic granulomatous disease, Chédiak-Higashi syndrome, and various disorders of chemotaxis, such as the lazy leukocyte syndrome. The most important factors predisposing to staphylococcal infections are not immunologic defects but mechanical defects. Minute skin abrasions, for example, probably provide the portal of entry in staphylococcal skin infections and in many cases of staphylococcal bacteremia. I.V. drug abuse accounts for many cases of staphylococcal bacteremia and endocarditis. Indwelling venous catheters are particularly important in nosocomial infections; plastic catheters become coated with fibrinogen and fibrin, which interact with adhesins on the bacterial cell surface and bind staphylococci to the catheter.

Clinical Staphylococcal Infections

Skin and soft tissue infections

The most frequent manifestations of staphylococcal disease are skin and soft tissue infections. These range from processes that can cause great discomfort but are rarely hazardous, such as impetigo, folliculitis, furuncles, carbuncles, and paronychia, to much more serious deep tissue infections, such as cellulitis and wound sepsis [see 2:VII Fungal, Bacterial, and Viral Infections of the Skin].

Staphylococcal skin infections are epidemiologically significant in the transmission of serious staphylococcal disease. Even localized infections can give rise to bacteremia and endocarditis, and furuncles of the nose and face may occasionally produce CNS infection such as cavernous vein thrombophlebitis by spreading along venous channels that lack valves.

Bone and joint infections

Staphylococci are among the principal causes of septic arthritis and osteomyelitis of the long bones and of vertebral bodies and disk spaces [see 7:XV Septic Arthritis and 7:XVI Osteomyelitis]. These infections may result from hematogenous seeding of proximal and distal long bones and joint spaces or contiguous spread of infection; infections of orthopedic prostheses are particularly serious.68 In addition to causing acute osteomyelitis, staphylococci have a propensity for causing chronic bone infections, usually of the metaphysis of long bones, in which active localized infection can persist for many decades without undergoing dissemination. S. aureus is also the leading cause of septic bursitis.

Respiratory tract infections

Although staphylococci reside in the nasopharynx, they do not cause pharyngitis; they are rarely involved in acute otitis, sinusitis, or mastoiditis, but they can participate in chronic infections of these regions [see 7:XIX Bacterial Infections of the Upper Respiratory Tract].

  1. aureusaccounts for fewer than 10% of all cases of bacterial pneumonia, but it is nevertheless an important pathogen of the lower respiratory tract. Staphylococcal pneumonia can occur either by spread from the upper airways or by hematogenous seeding. Except in infants, airborne staphylococcal pneumonia is rare as a primary infection. It usually occurs after viral respiratory tract infections, especially influenza. Thus, clinical illness often follows a biphasic course: viral symptoms lasting 4 or 5 days precede an abrupt deterioration as staphylococcal pneumonia progresses. Staphylococcal pneumonia also occurs as a nosocomial infection, particularly in patients on ventilators and in elderly, debilitated patients69; in such patients, mortality is 32%. S. aureuscan cause pneumonia in HIV-infected persons; these infections are often community acquired and have a mortality of 38%, even with appropriate antibiotic therapy.

Patients with staphylococcal pneumonia are acutely ill; purulent sputum production is the rule, and polymorphonuclear leukocytes and staphylococci can be identified on Gram stain of the sputum. Radiographs show a characteristic patchy bronchopneumonic pattern, often with multiple areas of consolidation. Because these organisms tend to produce tissue necrosis, cavitation is relatively common, and lung abscesses and empyemas may develop; fibrosis and scarring often occur during healing. Hematogenous staphylococcal pneumonia is a complication of staphylococcal bacteremia, especially when endocarditis of the tricuspid valve leads to septic embolization. Sputum production tends to occur later and is less prominent than in airborne pneumonias. Typical radiographic features include multiple small areas of infiltration, which, although sometimes evanescent, often progress to cavitation. Staphylococcal pneumonia is a life-threatening disease requiring high-dose parenteral antibiotic therapy, often for 2 to 4 weeks. Patients with abscess formation and empyema generally require therapy for 3 to 4 weeks; empyemas must be drained.

Bloodstream infections

Most patients with S. aureus bacteremia are acutely ill; more than 50% have temperatures in excess of 40° C (104° F), and most experience chills and exhibit systemic toxicity. Serious underlying diseases are present in 75% to 85% of patients. The overall mortality is about 23%.70 Up to 80% of patients with staphylococcal bacteremia acquire the infection in the hospital, often from infected I.V. catheters, skin and wound infections, or pulmonary tract infections.71 I.V. drug abuse is responsible for more than 50% of the community-acquired cases. Patients with community-acquired bacteremias are more likely to have endocarditis and secondary metastatic infections than patients with nosocomial infections, who are more likely to have an evident portal of entry and severe underlying diseases. Bacteremic seeding can lead to secondary staphylococcal infection of the lungs, bones and joints, and the genitourinary tract. Evidence of CNS involvement suggests endocarditis and is an adverse prognostic sign.

Patients with staphylococcal bacteremia have generally been treated with parenteral antibiotics for 4 to 6 weeks. This recommendation was made on the basis of the results of a study conducted from 1940 to 1954 in which endocarditis developed in 64% of 55 patients with staphylococcal bacteremia.72 Subsequent investigations suggested that patients with a removable or treatable primary focus of infection (most often, infected I.V. devices) who have no clinical evidence of endocarditis can be treated safely with only 10 to 21 days of I.V. antibiotics.73 Other studies, however, reported development of endocarditis in 22% to 38% of patients with a primary focus of bacteremia. Therefore, the possibility of endocarditis should not be ignored, even if a primary focus is present. Moreover, major complications, including shock, acute respiratory distress syndrome, and metastatic infection, can occur even in the absence of endocarditis.

Controlled trials are necessary to determine the safety and efficacy of short-term therapy for staphylococcal bacteremia. It may therefore be prudent to treat patients with staphylococcal bacteremia as though they have endocarditis, unless all of the following features are present: clear evidence of only a transient bacteremia; a removable primary focus of infection; a benign clinical course; an absence of echocardiographically demonstrable valvular abnormalities and metastatic infection; and intact host defenses. Negative titers of teichoic acid antibodies would bolster the decision to shorten the I.V. antibiotic therapy, which should, in any case, be continued for at least 10 to 15 days. A regimen of I.V. antibiotics followed by oral antibiotic therapy has been suggested, but more studies are needed before this regimen can be widely recommended. Clearly, the physician must individualize therapy by balancing the morbidity and expense of 4 to 6 weeks of antibiotic therapy against the risks of undertreating endocarditis, which is a life-threatening infection.74 Transesophageal echocardiography can establish a diagnosis of S. aureus endocarditis75 [see 7:XVIII Infective Endocarditis].

Because of the high mortality associated with staphylococcal bacteremia and endocarditis, combination therapies utilizing nafcillin or vancomycin with gentamicin or rifampin are being studied. Thus far, combination therapy appears to reduce the duration of bacteremia, but it does not change the long-term mortality. Combination treatment may also be useful in patients who fail to respond to conventional single-drug treatment. I.V. drug abusers with uncomplicated right-sided endocarditis may respond well to abbreviated therapy.

Because of the propensity of staphylococci to form abscesses, it is particularly important to evaluate bacteremic patients for a loculated infection that may require surgical drainage. Management of staphylococcal bacteremia in patients with Hickman catheters almost always requires catheter removal in addition to antibiotic therapy. Staphylococcal endocarditis appears to be increasing in frequency; its clinical features and therapy are reviewed elsewhere76 [see 7:XVIII Infective Endocarditis].

CNS infection

Staphylococcal infections of the CNS present most often as aseptic meningitis in patients with staphylococcal bacteremia, and especially those with endocarditis; the CSF typically reveals modest pleocytosis, and cultures and Gram stains are negative. S. aureus is an uncommon cause of purulent meningitis. Direct extension of infection from traumatic or neurosurgical wounds or from osteomyelitis of the mastoids or other cranial bones is a more frequent cause of staphylococcal meningitis than hematogenous seeding. Nafcillin has been shown to penetrate the CSF and should be administered in dosages of approximately 2 g I.V. every 4 hours for the average-sized adult with normal renal function. When nafcillin cannot be used because the infecting organisms are methicillin-resistant77 or because the patient is allergic to penicillin, therapy must be individualized. Options such as high-dose I.V. vancomycin and rifampin should be considered. In patients who do not respond favorably, there may be a role for intrathecal vancomycin or bacitracin; or for I.V. TMP-SMX, erythromycin, or chloramphenicol.

Another uncommon infection of the CNS in which S. aureus is a major pathogen is spinal epidural abscess. Patients with this disease often have underlying vertebral osteomyelitis. These patients present with fever and back pain; radicular pain, weakness, and paralysis evolve as cord compression occurs. Immediate surgical decompression is mandatory.

GI tract infections

Staphylococci can also affect the GI tract. Staphylococcal food poisoning occurs when food handlers who have contaminated superficial wounds or who are shedding infected nasal droplets inoculate foods with enterotoxin-producing strains of S. aureus. If the contaminated food is not refrigerated, the organisms produce enterotoxin within 4 to 6 hours. The symptoms begin abruptly within 1 to 6 hours after the ingestion of preformed enterotoxin. They consist of salivation, nausea, vomiting, abdominal pain, diarrhea, and prostration. Nausea and vomiting are the most prominent features, because of the enterotoxin's ability to affect the vomiting center of the brain. The symptoms usually subside within 24 hours; there is no specific therapy other than fluid and electrolyte repletion, although I.V. hydration may be necessary in cases of severe disease. In the 1950s and 1960s, GI superinfection or enterocolitis occurred in patients receiving tetracycline or chloramphenicol treatment. Recently, similar cases have occurred, caused by methicillin-resistant strains.

Genitourinary infections

  1. aureuscan occasionally produce genitourinary infections. From 7% to 27% of patients with staphylococcal bacteremia have positive urine cultures. Most of these patients have no gross renal infection, but in some, hematogenous pyelonephritis, renal cortical abscesses, renal carbuncles, perinephric abscesses, or prostatic abscesses may develop. These processes should be considered in all patients with positive urine cultures and in patients with staphylococcal bacteremia and persistent fever, flank pain, or abnormal urine sediments. I.V. pyelography, nephrotomography, renal ultrasonography, and CT may be diagnostically helpful. Prostatic abscesses may require repeated rectal examination or ultrasound for diagnosis. S. aureus, either alone or with other genitourinary pathogens, can also cause nosocomial urinary tract infections in patients with indwelling urethral catheters.

Other staphylococcal infections

  1. aureuscan produce infection of sebaceous glands of the eyelid (sty), purulent conjunctivitis, and orbital cellulitis. It can cause parotitis, which occurs most often in patients who are elderly, debilitated, and dehydrated. Decades ago, parotitis was most common in patients with renal failure, whereas today, it is most common in patients with Sjögren syndrome. Mastitis is most frequent in nursing mothers in the early postpartum period but may occur in neonates or in nonlactating women. Infection of muscle is rare in temperate climates but common in the tropics. Tropical pyomyositis presents as fever and multiple muscle abscesses. The diagnosis should be suspected in recent immigrants from tropical areas but should also be considered in patients who have not traveled and in patients with AIDS. CT may be helpful in detecting muscle abscesses. Purulent pericarditis can result either from direct extension of a pneumonia or empyema or from hematogenous seeding. Staphylococcal bacteremia can produce metastatic abscesses of any organ, including the liver and spleen. Patients who have metastatic abscesses may have persistent fever; in such patients, diagnosis may be difficult if localizing symptoms are absent. S. aureuscan cause peritonitis in patients on continuous ambulatory peritoneal dialysis.

Treatment of Staphylococcal Infections

It is extremely important to establish a microbiologic diagnosis in staphylococcal infections and to utilize cultures and sensitivities to direct therapy. Choice of empirical antibiotics and of the route of administration depend on the seriousness of the infection, the trends in staphylococcal resistance patterns in the specific geographical area, and the risk factors for methicillin resistance in a given patient.

In addition to paronychia and sties, minor staphylococcal infections of the skin, such as folliculitis and furunculosis, generally respond well to the topical application of warm soaks. Larger focal infections such as carbuncles may have to be incised and drained. Antibiotics should be added when fever or systemic symptoms are present, when lesions are large or numerous, when lesions fail to respond to local therapy, when patients have underlying medical problems such as valvular heart disease or a cardiac prosthesis, or when the nose or face is involved. Cloxacillin or dicloxacillin in an oral dosage of 250 to 500 mg every 6 hours is generally sufficient. Cephalexin, erythromycin, or the newer macrolides are excellent alternatives, as is clindamycin in an oral dosage of 150 to 300 mg every 6 hours [see Table 9].

Table 9 Antibiotic Treatment for Staphylococcal Infections






Cloxacillin or

250–500 mg q. 6 hr

Excellent efficacy


250 mg q. 6 hr

Alternative for penicillin-allergic patients with cellulitis, pneumonia, or osteomyelitis


150–300 mg q. 6 hr

Alternative for penicillin-allergic patients with cellulitis, pneumonia, or osteomyelitis


600 mg q.d.

250 mg b.i.d.

May be useful in right-sided endocarditis; utilized because of noncompliance in I.V. drug abusers

Oral or


600 mg q. 12 hr

For MRSA; may be more suitable than other anti-MRSA agents for staphylococcal TSS; for vancomycin-resistant infection



1 g q. 12 hr

Alternative for penicillin-allergic patients with endocarditis and for MRSA

Oxacillin or nafcillin

For cellulitis or pneumonia:
  1 g q. 4 hr
For osteomyelitis, endocarditis,
  CNS infections:
  2 g q. 4 hr

Favored choice for non-MRSA strains


Penicillin G

2 g I.V. q. 6 hr
1,500 mg I.V. q. 8 hr

2–4 million units
  q. 4–6 hr

Alternative agent unless CNS infection is present

First choice for sensitive organisms


4 mg/kg/day

For MRSA, except in endocarditis; for vancomycin-resistant infection; dose-related


7.5 mg/kg q. 8 hr

For MRSA; can cause phlebitis and myopathy; requires central line placement

CNS—central nervous system  MRSA—methicillin-resistant Staphylococcus aureus  TSS—toxic-shock syndrome

For serious staphylococcal infections, such as cellulitis, deep wound sepsis, pneumonia, septic arthritis, osteomyelitis, bacteremia, endocarditis, or meningitis, parenteral antibiotics are mandatory. Oxacillin or nafcillin in a dosage of 1 g every 4 hours may suffice for cellulitis or pneumonia [see 7:XIV Chemotherapy of Infection], but a dosage of 2 g every 4 hours should be used for osteomyelitis, endocarditis, or CNS infections in the average-sized adult with normal renal function. Parenteral cephalosporins such as cephalothin and cefuroxime are excellent alternatives unless CNS infection is present. For patients who are allergic to penicillin, erythromycin and clindamycin have produced good results in the treatment of cellulitis, pneumonia, and osteomyelitis; vancomycin is preferred for the treatment of endocarditis in these patients because it is a bactericidal antibiotic.

Although only about 10% of all S. aureus strains are penicillin sensitive, penicillin G remains the drug of choice for sensitive organisms. Care must be taken in sensitivity testing, however, because penicillinase is an inducible enzyme. It is best to reserve penicillin G for staphylococci that are shown to be penicillinase negative or for staphylococci that have extremely large zones of inhibition by penicillin on Kirby-Bauer sensitivity testing. Methicillin resistance may be overlooked occasionally on routine sensitivity testing because it is more readily expressed at lower temperatures (30° C), at higher salt concentrations, or after 48 hours of incubation.

Synergism between penicillins and aminoglycosides against S. aureus has been demonstrated in vitro and in experimental staphylococcal infections. However, a clinical benefit of combination therapy for S. aureus endocarditis has not yet been demonstrated. Most staphylococci are extremely sensitive to rifampin, which may be combined with other antistaphylococcal drugs in certain difficult clinical situations. However, staphylococci rapidly develop resistance to rifampin when the drug is used alone, and resistance may even arise during combination therapy. The utility of ciprofloxacin is limited by the emergence of resistance. Only time will tell if the newer fluoroquinolones, such as levofloxacin and sparfloxacin, will be more successful; combinations of rifampin and a fluoroquinolone are being investigated.


Since its first appearance in 1970, MRSA has been a growing problem. These organisms produce a cell-wall penicillin-binding protein with a low affinity for β-lactam antibiotics. Initially, MRSA was a nosocomial pathogen in university hospitals. Although hospitals are still the most common setting, these organisms have also become important problems in long-term care facilities and are increasingly being acquired in the community.78 MRSA has proved to be difficult to control; infected patients should be subjected to strict barrier precautions.79 Both patients and hospital personnel may become asymptomatic carriers. The carrier state is typically prolonged, often exceeding 3 years; antibiotics are usually ineffective. Topical mupirocin ointment can reduce the MRSA carrier rate,80 but because recolonization is common, mupirocin is not recommended for extended use in long-term care facilities.

The virulence and clinical manifestations of MRSA are no different from those of methicillin-susceptible S. aureus81; compared to methicillin-sensitive strains, however, a higher percentage of MRSA strains possess the toxins TSST-1, enterotoxins, and the Panton-Valentine leukocidin. MRSA strains are also resistant to all β-lactam antibiotics, erythromycin, and chloramphenicol; even if the organisms appear to be sensitive to cephalosporins in disk diffusion testing, one should not rely on these agents. In the past, vancomycin has been the only alternative for treating MRSA. Interestingly, vancomycin is less effective than nafcillin for strains sensitive to both agents. Linezolid, daptomycin, teicoplanin, and quinupristin-dalfopristin are newer antibiotics with activity against MRSA. Daptomycin is approved for skin and soft tissue infection, but because it is rapidly bactericidal, it needs to be studied as a treatment for endocarditis caused by MRSA. Linezolid has been approved for skin and soft tissue infections, as well as for pneumonia caused by MRSA; evidence is accumulating that, like clindamycin, it is a potent suppressor of staphylococcal toxin production. Because of this characteristic, it may be a more suitable agent to use in patients with staphylococcal TSS. Quinupristin-dalfopristin often causes phlebitis and myopathy. Thus, its use requires placement of a central line.

Vancomycin resistance

With the increased use of vancomycin, strains of S. aureus with reduced susceptibility to vancomycin have begun to appear.67 Although these organisms are still very uncommon in the United States, vigorous control measures are required to prevent them from joining VRE as major nosocomial pathogens.82 Daptomycin and linezolid have excellent activity against vancomycin-intermediate and vancomycin-resistantStaphylococcus.

Adjunctive measures

In addition to vigorous antibiotic therapy, other measures are often necessary to treat staphylococcal infections. In particular, it is important to remove indwelling venous catheters or other foreign bodies that may be a portal of entry for staphylococcal bacteremia or a nidus for persistent infection. Patients must be evaluated for the presence of staphylococcal abscesses, which often must be drained. Finally, endocarditis and metastatic infection are a concern in all patients with staphylococcal bacteremia.


Epidemiologic control of staphylococcal infection requires ongoing surveillance and the reporting of infections. Contact precautions should be followed in the management of patients with active infections of the skin or wounds. Respiratory precautions may be useful in the management of patients with staphylococcal pneumonia, although droplet spread is much less important than transmission by direct contact. The treatment of nasal or rectal carriers can be frustrating, particularly if the carriers are hospital personnel or persons suffering from recurrent furunculosis. Topical treatment with germicidal soaps, povidone-iodine, or antibiotic ointments has been advocated, but long-term results have been disappointing. Orally administered antibiotics, including rifampin, TMP-SMX, and ciprofloxacin, have also failed to live up to initially promising findings. Bacterial interference, which attempts to replace epidemiologically virulent strains of staphylococci with strains that have been deliberately colonized and are less virulent, has generally been abandoned, in part because infection has been caused by these supposedly less virulent strains.

Attempts to develop staphylococcal vaccines are continuing.83 In a clinical trial in dialysis patients, vaccination with a staphylococcal surface carbohydrate conjugated to Pseudomonas exotoxin A significantly reduced the incidence of bacteremia, though protective antibodies lasted only 8 months.84

Intranasal therapy with mupirocin ointment may help control the staphylococcal carrier state. A naturally occurring antibiotic produced by P. fluorescens, mupirocin inhibits S. aureus (including penicillin- and methicillin-resistant strains) and other aerobic gram-positive cocci by binding reversibly to bacterial isoleucyl transfer RNA synthetase; no cross-resistance has been observed between mupirocin and other antibiotics. A placebo-controlled trial of mupirocin in 34 patients who were S. aureus carriers found that a monthly course of nasal mupirocin reduced the incidence of nasal colonization and skin infections for at least 1 year. Mupirocin was administered intranasally twice daily for 5 days; side effects were minimal.85 Additional studies have confirmed the benefits of mupirocin. However, because resistance to mupirocin and recolonization after therapy can occur, routine use of mupirocin should be avoided.80 The FDA has approved mupirocin for the topical treatment of impetigo.

Staphylococcal Toxic-Shock Syndrome

Staphylococcal TSS was first reported in 1978; by 1990, more than 3,300 cases had been reported in the United States, 90% of which occurred during menstruation in women who were using tampons. The incidence of staphylococcal TSS declined precipitously after superabsorbent tampons were withdrawn from the market. Currently, fewer than 100 cases occur each year, and nonmenstrual cases occur more often than those associated with menstruation. Most cases are now nosocomially acquired, often as a result of postoperative staphylococcal wound infections, particularly those associated with nasal packing after rhinoplasty. Staphylococcal TSS caused by toxin-producing strains of MRSA has been reported from the United States and Japan.

TSS is a multisystem disease with diverse clinical manifestations [see Table 10]. Leukocytosis and thrombocytopenia (< 100,000 platelets/mm3) are common findings. Urinalysis may show mild pyuria and, occasionally, microscopic hematuria. Blood urea nitrogen and creatinine levels are elevated in more than 50% of patients. Serum bilirubin and hepatic enzyme levels are elevated in about half of patients. Serum creatine kinase levels are high in more than one third of patients, and myoglobinuria has developed in some patients. Elevated serum amylase levels are also found, but such elevations may be related to azotemia rather than to clinically evident pancreatitis. Unexplained marked hypocalcemia is often observed. The drop in serum calcium level is out of proportion to the degree of hypoalbuminuria noted in some patients and may be caused by elevated serum calcitonin levels. Blood cultures show no growth in almost all cases. Group A streptococci can produce a severe form of TSS that resembles staphylococcal TSS; in streptococcal TSS, however, bacteremia is common, necrotizing fasciitis may be present, and mortality is much higher (30% to 70%) [see Streptococcal Toxic-Shock Syndrome, above].

Table 10 Clinical Manifestations of Staphylococcal Toxic-Shock Syndrome

High fever
Scarlatiniform eruption
Hypotension and shock
Desquamation, particularly of palms and soles, during convalescence
  Mucous membranes: hyperemia of conjunctivae, pharynx, and vagina
  Gastrointestinal tract: vomiting and diarrhea at onset
  Muscle: severe myalgias (elevated serum creatine kinase level; rhabdomyolysis in some patients)
  Central nervous system: toxic encephalopathy (disorientation, delirium, or obtundation in absence of high fever or shock)
  Kidney: azotemia; pyuria in absence of demonstrable urinary tract infection
  Liver: elevations of serum bilirubin and aspartate aminotransferase
  Blood: thrombocytopenia


The management of staphylococcal TSS calls for immediate correction of hypotension and shock with vigorous fluid replacement, attention to the site of S. aureus colonization or infection (e.g., removal of tampon and drainage of any abscess), and systemic antimicrobial therapy with an antistaphylococcal agent. Albumin replacement may be needed to counter capillary leak; in addition, correction of hypocalcemia, use of renal dialysis, and ventilator support may be necessary. Because of the emergence of staphylococcal TSS caused by MRSA, empirical antibiotic therapy should be with vancomycin, daptomycin, or linezolid [see Table 11]. Protein synthesis inhibitors such as clindamycin and linezolid have been shown to suppress TSST-1 production, so linezolid is a reasonable choice. In contrast, cell-wall-active agents cause release of TSST-1. A retrospective analysis of 45 patients suggested that glucocorticoids may assist in recovery,86 but more data are needed before glucocorticoids can be recommended for all patients with staphylococcal TSS. Most patients recover in 1 to 2 weeks; mortality is about 5%.

Table 11 Antibiotic Treatment for Staphylococcal Toxic-Shock Syndrome*





1 g I.V. q. 12 hr

Consider if patient has risk factors for MRSA


4 mg/kg/day

Rapidly bactericidal


600 mg I.V. q. 12 hr

Suppresses toxin synthesis; has good activity against MRSA


600–800 mg I.V. q. 8 hr

Suppresses toxin synthesis; inducible clindamycin resistance has been described in community-acquired MRSA


2 g I.V. q. 4 hr

For non-MRSA infection

*No comprehensive studies of antibiotic treatment in staphylococcal toxic shock syndrome have been done. In general, clindamycin and linezolid are favored, because of their ability to suppress toxin synthesis; conversely, cell-wall-active agents cause release of toxin.


  1. epidermidisand other coagulase-negative staphylococcal species are organisms of low pathogenic potential that are universally present as part of the normal skin flora. Because of their ubiquity, coagulase-negative staphylococci are frequently isolated from clinical specimens, including blood cultures. Most often, they are skin contaminants rather than true pathogens; unfortunately, neither clonal variability87nor molecular typing88 can reliably distinguish between contaminants and pathogens. In certain circumstances, however, these organisms can cause significant disease. In the great majority of cases, coagulase-negative staphylococci are opportunistic pathogens, producing infection only in patients with significant underlying medical problems.89 They most often produce infections in patients with indwelling I.V. catheters or implanted prosthetic devices. Many strains produce an exopolysaccharide that functions as a capsule.90 This slime layer is an adhesive that binds the organism to implanted devices; it also inhibits the action of antibiotics, further contributing to pathogenicity.

Coagulase-negative staphylococci are now the leading cause of nosocomial bacteremia, which most often results from the use of I.V. catheters, especially centrally placed catheters. Patients with this infection may have signs of phlebitis at the needle or catheter site and may present with persistent low-grade fevers or high-spiking temperature elevations. The I.V. needle or catheter should be removed and a course of parenteral antibiotics administered. Metastatic infection is uncommon but can be serious. These organisms can also be an important cause of bacteremia in immunosuppressed patients. The mortality in patients with coagulase-negative staphylococcal bacteremia approaches 40%, in part because these patients have such serious underlying diseases; mortality attributable to the infection itself is about 13%.

An even more serious problem occurs when S. epidermidis or other coagulase-negative staphylococci infect indwelling ventriculoatrial shunts. Patients with this condition may present with bacteremia, meningitis, or both. In addition to high-dose antibiotic therapy, management must often include shunt removal or revision. Vascular grafts and joint prostheses may also become infected. These infections are generally indolent and are more likely to produce pain and joint dysfunction than fever and local inflammatory signs. Diagnosis may be difficult because coagulase-negative staphylococci can be recovered from joint aspirates, either as pathogens or contaminants. These organisms can sometimes cause indolent wound infections or osteomyelitis. Here, too, diagnosis may be difficult, but visualization of the staphylococci on Gram stain, their consistent isolation on culture, and the absence of other pathogens should suggest their etiologic role.

The most serious therapeutic problem caused by coagulase-negative staphylococci is prosthetic valve endocarditis. The disease typically has a subacute course, but eradication of the organisms is often very difficult because of antibiotic resistance. Medical therapy with high-dose antibiotics should be attempted, but valve replacement is necessary if infection persists or if significant valve dysfunction occurs. Mortality is high, approaching 50% with or without surgical therapy. Coagulase-negative staphylococci may also cause endocarditis on native valves.89

Antibiotic therapy for deep-seated coagulase-negative staphylococcal infections is difficult. Fifty percent of strains are resistant to methicillin and other semisynthetic penicillins. Whereas most strains appear to be sensitive to cephalosporins on disk diffusion testing, these results correlate poorly with actual bactericidal activity. Vancomycin, gentamicin, and rifampin are bactericidal against most coagulase-negative staphylococci. Vancomycin or rifampin resistance is rare but may emerge during therapy. Vancomycin is generally the drug of choice, but for serious infections, the addition of rifampin or gentamicin should be strongly considered. Linezolid and quinupristin-dalfopristin have good activity against methicillin-resistant coagulase-negative staphylococci, and clinical trials have proved their efficacy in vivo.91 Although these agents are good alternatives, additional studies are clearly needed.

  1. saprophyticusis a coagulase-negative member of the normal skin flora that can be distinguished from S. epidermidisby its resistance to the antibiotic novobiocin. Previously dismissed as a nonpathogen, S. saprophyticus is now clearly recognized to be an important cause of urinary tract infections. S. saprophyticus is second only to E. coli as a cause of cystitis in sexually active young women, accounting for 10% to 15% of such cases.92 S. saprophyticus is much less likely to produce infections in men; however, S. saprophyticus commonly infects elderly men who have undergone urinary tract instrumentation. The clinical features of S. saprophyticus infections resemble those of other urinary tract infections; bacteremia and systemic toxicity are rare, even when upper urinary tract infection occurs. S. saprophyticus is sensitive to most antimicrobial agents that are used to treat urinary tract infections.


  1. Gillespie SH, Balakrishnan I: Pathogenesis of pneumococcal infection. J Med Microbiol 49:1057, 2000
  2. Obaro SK, Monteil MA, Henderson DC: The pneumococcal problem. BMJ 312:1521, 1996
  3. Nuroti JP, Butler JC, Farley MM, et al: Cigarette smoking and invasive pneumococcal disease. N Engl J Med 342:681, 2000
  4. Kyaw MH, Christie P, Clarke SC, et al: Invasive pneumococcal disease in Scotland, 1999–2001: use of record linkage to explore associations between patients and disease in relation to future vaccination policy. Clin Infect Dis 37:1283, 2003
  5. Feikin DR, Schuchat A, Kolczak M, et al: Mortality from invasive pneumococcal pneumonia in the era of antibiotic resistance, 1995–1997. Am J Public Health 90:223, 2000
  6. Mufson MA, Stanek RJ: Bacteremic pneumococcal pneumonia in one American city: a 20-year longitudinal study, 1978–1997. Am J Med 107(1A):34S, 1999
  7. Fiore AE, Moroney JF, Farley MM, et al: Clinical outcomes of meningitis caused by Streptococcus pneumoniaein the era of antibiotic resistance. Clin Infect Dis 30:71, 2000
  8. Rubins JB, Cheung S, Carson P, et al: Identification of clinical risk factors for nosocomial pneumococcal bacteremia. Clin Infect Dis 29:178, 1999
  9. O'Brien KL, Walters MI, Sellman J, et al: Severe pneumococcal pneumonia in previously healthy children: the role of preceding influenza infection. Clin Infect Dis 30:784, 2000
  10. Osler W: The Principles and Practice of Medicine: Designed for the Use of Practitioners and Students of Medicine. D Appleton & Co, New York, 1892, p 517
  11. Dowell SF, Garman RL, Liu G, et al: Evaluation of binax NOW, an assay for the detection of pneumococcal antigen in urine samples, performed among pediatric patients. Clin Infect Dis 32:824, 2001
  12. Harrison LH, Dwyer DM, Billmann L, et al: Invasive pneumococcal infection in Baltimore, MD: implications for immunization policy. Arch Intern Med 160:89, 2000
  13. Stevens DL, Rout RK, Rosen H, et al: Differential white blood cell count: underutilized or overrated? Curr Opin Infect Dis 13:205, 2000
  14. Breiman RF, Keller DW, Phelan MA, et al: Evaluation of effectiveness of the 23-valent pneumococcal capsular polysaccharide vaccine for HIV-infected patients. Arch Intern Med 160:2633, 2000
  15. Siegel M, Timpone J: Penicillin-resistant Streptococcus pneumoniaeendocarditis: a case report and review. Clin Infect Dis 32:972, 2001
  16. Tomasz A: New faces of an old pathogen: emergence and spread of multidrug-resistant Streptococcus pneumoniae. Am J Med 107(1A):55S, 1999
  17. Whitney CG, Farley M, Hadler J, et al: Increasing prevalence of multidrug-resistant Streptococcus pneumoniaein the United States. N Engl J Med 343:1917, 2000
  18. Metlay JP, Hofmann J, Cetron MS, et al: Impact of penicillin susceptibility on medical outcomes for adult patients with bacteremic pneumococcal pneumonia. Clin Infect Dis 30:520, 2000
  19. Normark BH, Novak R, Ortqvist A, et al: Clinical isolates of Streptococcus pneumoniaethat exhibit tolerance of vancomycin. Clin Infect Dis 32:552, 2001
  20. Ho PL, Tse WS, Tsang KWT, et al: Risk factors for acquisition of levofloxacin-resistant Streptococcus pneumoniae: a case-control study. Clin Infect Dis 32:701, 2001
  21. Plouffe JF: Emerging therapies for serious gram-positive bacterial infections: a focus on linezolid. Clin Infect Dis 31(suppl 4):S144, 2000
  22. Ball P: Therapy for pneumococcal infection at the millennium: doubts and certainties. Am J Med 107(1A):77S, 1999
  23. Ramirez JA, Bordon J: Early switch from intravenous to oral antibiotics in hospitalized patients with bacteremic community-acquiredStreptococcus pneumoniaepneumonia. Arch Intern Med 161:848, 2001
  24. Recommendations of the Advisory Committee on Immunization Practices (ACIP): Preventing pneumococcal disease among infants and young children. MMWR Morb Mortal Wkly Rep 49:1, 2000
  25. Blumberg EA, Brozena SC, Stutman P, et al: Immunogenicity of pneumococcal vaccine in heart transplant recipients. Clin Infect Dis 32:307, 2001
  26. Recommendations of the Advisory Committee on Immunization Practices (ACIP): Prevention of pneumococcal disease. MMWR Morb Mortal Wkly Rep 46:1, 1997
  27. Jackson LA, Neuzil KM, Yu O, et al: Effectiveness of pneumococcal polysaccharide vaccine in older adults. N Engl J Med 348:1747, 2003
  28. Ament A, Baltussen R, Duru G, et al: Cost-effectiveness of pneumococcal vaccination of older people: a study of 5 Western European countries. Clin Infect Dis 31:444, 2000
  29. Whitney CG, Farley MM, Hadler J, et al: Decline in invasive pneumococcal disease after the introduction of protein-conjugate vaccine. N Engl J Med 348:1737, 2003
  30. Mufson MA: Antibody response of pneumococcal vaccine: need for booster dosing? Int J Antimicrob Agents 14:107, 2000
  31. Jackson LA, Benson P, Sneller V-P, et al: Safety of revaccination with pneumococcal polysaccharide vaccine. JAMA 281:243, 1999
  32. Carapetis JR, Currie BJ, Kaplan EL: Epidemiology and prevention of group A streptococcal infections: acute respiratory tract infections, skin infections, and their sequelae at the close of the twentieth century. Clin Infect Dis 28:205, 1999
  33. O'Brien KL, Beall B, Barrett NL, et al: Epidemiology of invasive group A Streptococcusdisease in the United States, 1995–1999. Clin Infect Dis 35:268, 2002
  34. Bisno AL, Gerber MA, Gwaltney JM, et al: Diagnosis and management of group A streptococcal pharyngitis: a practice guideline. Clin Infect Dis 25:574, 1997
  35. De Quirós JCLB, Moreno S, Cercenado E, et al: Group A streptococcal bacteremia: a 10-year prospective study. Medicine (Baltimore) 76:238, 1997
  36. Baddour LM: Infective endocarditis caused by β-hemolytic streptococci. The Infectious Diseases Society of America's Emerging Infections Network. Clin Infect Dis 26:66, 1998
  37. Sommer R, Rohner P, Garbino J, et al: Group A β-hemolytic Streptococcusmeningitis: clinical and microbiological features of nine cases. Clin Infect Dis 29:929, 1999
  38. Dupuy A, Benchikhi H, Roujeau J-C, et al: Risk factors for erysipelas of the leg (cellulitis): case-control study. BMJ 318:1591, 1999
  39. Bisno AL, Stevens DL: Streptococcal infections of skin and soft tissues. N Engl J Med 334:240, 1996
  40. Kaul R, McGeer A, Low DE, et al: Population-based surveillance for group A streptococcal necrotizing fasciitis: clinical features, prognostic indicators, and microbiologic analysis of seventy-seven cases. Am J Med 103:18, 1997
  41. Zurawski CA, Bardsley M, Beall B, et al: Invasive group A streptococcal disease in metropolitan Atlanta: a population-based assessment. Clin Infect Dis 27:150, 1998
  42. Stevens DL: Streptococcal toxic shock syndrome. Clin Microbiol Infect 8:133, 2002
  43. Stevens DL: Group A streptococcal sepsis. Curr Infect Dis Rep 5:379, 2003
  44. Darenberg OJ, Ihendyane N, Sjolin J, et al: Intravenous immunoglobulin G therapy in streptococcal toxic shock syndrome: a European randomized, double-blind, placebo-controlled trial. Clin Infect Dis 37:333, 2003
  45. Stollerman GH: Rheumatic fever. Lancet 349:935, 1997
  46. Williamson L, Bowness P, Mowat A, et al: Lesson of the week: difficulties in diagnosing acute rheumatic fever—arthritis may be short lived and carditis silent. BMJ 320:362, 2000
  47. Guidelines for the diagnosis of rheumatic fever. Jones Criteria, 1992 update. Special Writing Group of the Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease of the Council on Cardiovascular Disease in the Young of the American Heart Association. JAMA 268:2069, 1992
  48. Adam D, Scholz H, Helmerking M: Short-course antibiotic treatment of 4782 culture-proven cases of group A streptococcal tonsillopharyngitis and incidence of poststreptococcal sequelae. J Infect Dis 182:509, 2000
  49. Lan AJ, Colford JM Jr: The impact of dosing frequency on the efficacy of 10-day penicillin or amoxicillin therapy for streptococcal tonsillopharyngitis: a meta-analysis. Pediatrics 105:E19, 2000
  50. Eriksson BKG: Anal colonization of group G β-hemolytic streptococci in relapsing erysipelas of the lower extremity. Clin Infect Dis 29:1319, 1999
  51. Hirose Y, Yagi K, Honda H, et al: Toxic shock-like syndrome caused by non-group A β-hemolytic streptococci. Arch Intern Med 157:1891, 1997
  52. Krohn MA, Hillier SL, Baker CJ: Maternal peripartum complications associated with vaginal group B streptococci colonization. J Infect Dis 179:1410, 1999
  53. Schrag SJ, Zywicki S, Farley MM, et al: Group B streptococcal disease in the era of intrapartum antibiotic prophylaxis. N Engl J Med 342:15, 2000
  54. Henning KJ, Hall EL, Dwyer DM, et al: Invasive group B streptococcal disease in Maryland nursing home residents. J Infect Dis 183:1138, 2001
  55. Gardam MA, Low DE, Saginur R, et al: Group B streptococcal necrotizing fasciitis and streptococcal toxic shock-like syndrome in adults. Arch Intern Med 158:1704, 1998
  56. Kupferwasser I, Darius H, Muller AM, et al: Clinical and morphological characteristics in Streptococcus bovisendocarditis: a comparison with other causative microorganisms in 177 cases. Heart 80:276, 1998
  57. Shenep JL: Viridans-group streptococcal infections in immunocompromised hosts. Int J Antimicrob Agents 14:129, 2000
  58. Cabellos C, Viladrich PF, Corredoira J, et al: Steptococcal meningitis in adult patients: current epidemiology and clinical spectrum. Clin Infect Dis 28:1104, 1999
  59. Bert F, Bariou-Lancelin M, Lambert-Zechovsky N: Clinical significance of bacteremia involving the “Streptococcus milleri” group: 51 cases and review. Clin Infect Dis 27:385, 1998
  60. Murray BE: Vancomycin-resistant enterococcal infections. N Engl J Med 342:710, 2000
  61. Donskey CJ, Chowdhry TK, Hecker MT, et al: Effect of antibiotic therapy on the density of vancomycin-resistant enterococci in the stool of colonized patients. N Engl J Med 343:1925, 2000
  62. Caballero-Granado FJ, Becerril B, Cuberos L, et al: Attributable mortality rate and duration of hospital stay associated with enterococcal bacteremia. Clin Infect Dis 32:587, 2001
  63. Winston DJ, Emmanouilides C, Kroeber A, et al: Quinupristin/dalfopristin therapy for infections due to vancomycin-resistantEnterococcus faecium. Clin Infect Dis 30:790, 2000
  64. Tenorio AL, Badri SM, Sahgal NB, et al: Effectiveness of gloves in the prevention of hand carriage of vancomycin-resistantEnterococcusspecies by health care workers after patient care. Clin Infect Dis 32:826, 2001
  65. Lowy FD: Staphylococcus aureusinfections. N Engl J Med 339:520, 1998
  66. von Eiff C, Becker K, Machka K, et al: Nasal carriage as a source of Staphylococcus aureusbacteremia. N Engl J Med 344:11, 2001
  67. Fridkin SK: Vancomycin-intermediate and -resistant Staphylococcus aureus: what the infectious disease specialist needs to know. Clin Infect Dis 32:108, 2001
  68. Murdoch DR, Roberts SA, Fowler VG Jr, et al: Infection of orthopedic prostheses after Staphylococcus aureusbacteremia. Clin Infect Dis 32:647, 2001
  69. Gonzalez C, Rubio M, Romero-Vivas J, et al: Bacteremic pneumonia due to Staphylococcus aureus: a comparison of disease caused by methicillin-resistant and methicillin-susceptible organisms. Clin Infect Dis 29:1171, 1999
  70. Mylotte MJ, Tayara A: Staphylococcus aureusbacteremia: predictors of 30-day mortality in a large cohort. Clin Infect Dis 31:1170, 2000
  71. Jensen AG, Wachmann CH, Poulsen KB, et al: Risk factors for hospital-acquired Staphylococcus aureusbacteremia. Arch Intern Med 159:1437, 1999
  72. Wilson R, Hamburger M: Fifteen years' experience with Staphylococcussepticemia in a large city hospital: analysis of fifty-five cases in the Cincinnati General Hospital, 1940 to 1954. Am J Med 22:437, 1957
  73. Nolan CM, Beaty HN: Staphylococcus aureusbacteremia: current clinical patterns. Am J Med 60:495, 1976
  74. Rader BL, Wandall DA, Frimodt-Moller N, et al: Clinical features of Staphylococcus aureusendocarditis: a 10-year experience in Denmark. Arch Intern Med 159:462, 1999
  75. Fowler VG Jr, Sanders LL, Kong LK, et al: Infective endocarditis due to Staphylococcus aureus: 59 prospectively identified cases with follow-up. Clin Infect Dis 28:106, 1999
  76. Tornos P, Almirante B, Mirabet S, et al: Infective endocarditis due to Staphylococcus aureus: deleterious effect of anticoagulant therapy. Arch Intern Med 159:473, 1999
  77. Lu C-H, Chang W-N: Adults with meningitis caused by oxacillin-resistant Staphylococcus aureus. Clin Infect Dis 31:723, 2000
  78. Gorak EJ, Yamada SM, Brown JD: Community-acquired methicillin-resistant Staphylococcus aureusin hospitalized adults and children without known risk factors. Clin Infect Dis 29:797, 1999
  79. Kotilainen P, Routamaa M, Peltonen R, et al: Eradication of methicillin-resistant Staphylococcus aureusfrom a health center ward and associated nursing home. Arch Intern Med 161:859, 2001
  80. Laupland KB, Conly JM: Treatment of Staphylococcus aureuscolonization and prophylaxis for infection with topical intranasal mupirocin: an evidence-based review. Clin Infect Dis 37:933, 2003
  81. Stevens DL, Herr D, Lampiris H, et al: Linezolid versus vancomycin for the treatment of methicillin-resistant Staphylococcus aureusinfections. Clin Infect Dis 34:1481, 2002
  82. Hiramatsu K: Vancomycin-resistant Staphylococcus aureus: a new model of antibiotic resistance. Lancet Infect Dis 1:147, 2001
  83. McKenney D, Poullot KL, Wang Y, et al: Broadly protective vaccine for Staphylococcus aureusbased on an in vivo-expressed antigen. Science 284:1523, 1999
  84. Shinefield H, Black S, Fattom A, et al: Use of a Staphylococcus aureusconjugate vaccine in patients receiving hemodialysis. N Engl J Med 346:491, 2002
  85. Raz R, Miron D, Colodner R, et al: A 1-year trial of nasal mupirocin in the prevention of recurrent staphylococcal nasal colonization and skin infection. Arch Intern Med 156:1109, 1996
  86. Todd JK, Ressman M, Caston SA, et al: Corticosteroid therapy for patients with toxic shock syndrome. JAMA 252:3399, 1985
  87. Van Eldere J, Peetermans WE, Struelens M, et al: Polyclonal staphylococcal endocarditis caused by genetic variability. Clin Infect Dis 31:24, 2000
  88. Seo SK, Venkataraman L, DeGirolami PC, et al: Molecular typing of coagulase-negative staphylococci from blood cultures does not correlate with clinical criteria for true bacteremia. Am J Med 109:697, 2000
  89. Proctor RA: Editorial response: coagulase-negative staphylococcal infections: a diagnostic and therapeutic challenge. Clin Infect Dis 31:31, 2000
  90. Fey PD, Ulphani JS, Gotz F, et al: Characterization of the relationship between polysaccharide intercellular adhesion and hemagglutination in Staphylococcus epidermidis. J Infect Dis 179:1561, 1999
  91. Clark NM, Hershberger E, Zervosc MJ, et al: Antimicrobial resistance among gram-positive organisms in the intensive care unit. Curr Opin Crit Care 9:403, 2003
  92. Ronald A: The etiology of urinary tract infection: traditional and emerging pathogens. Am J Med 113(suppl 1A):14S, 2002
  93. Moellering RC: Enterococcusspecies, Streptococcus bovis, and Leuconostoc species. Principles and Practice of Infectious Diseases, 5th ed. Mandell GL, Bennett JE, Dolin R, Eds. Churchill Livingstone, Philadelphia, 2000, p 2147
  94. Martin DR: Laboratory evaluation of streptococci. Streptococcal Infections: Clinical Aspects, Microbiology and Molecular Pathogenesis. Stevens DL, Kaplan EL, Eds. Oxford University Press, New York, 2000, p 266
  95. Bisno AL: Nonsuppurative poststreptococcal sequelae: rheumatic fever and glomerulonephritis. Principles and Practice of Infectious Diseases, 5th ed. Mandell GL, Bennett JE, Dolin R, Eds. Churchill Livingstone, Philadelphia, 2000, p 2117

Editors: Dale, David C.; Federman, Daniel D.