Streptococci

Streptococci are gram-positive cocci that, on Gram stain, typically occur in chains. Streptococci are subclassified in several ways. The three most useful classifications are by bacterial oxygen requirements, by colony appearance on blood agar, and by specific carbohydrate from the organism. Depending on clinical oxygen environment associated with disease, streptococci may be considered aerobic, microaerophilic, or anaerobic. Most streptococci are aerobic. The microaerophilic organisms sometimes cause a chronic resistant type of skin ulcer and occasionally are isolated in deep wound infections. The anaerobic streptococci are discussed later.

Streptococci are divided into three types by the appearance of the hemolysis that the streptococcal colonies produce on sheep blood agar—alpha, beta, and gamma. Alpha (hemolytic) streptococci are characterized by incomplete hemolysis surrounding the colony; this area usually has a greenish appearance on sheep blood agar, and streptococci producing green hemolysis are often called “viridans.” The beta hemolytic organisms produce a complete, clear (colorless) hemolysis. Gamma streptococci do not produce hemolysis on sheep blood agar. These differences have clinical value. Alpha streptococci of the viridans subgroup are one of the most frequent causes of subacute bacterial endocarditis. Beta streptococci are the causative agent of several different types of infection and syndromes, as will be discussed later, and account for the great majority of disease associated with streptococci with the exception of subacute bacterial endocarditis. Gamma streptococci are of lesser importance, but some of the organisms known as enterococci belong to this category.

The third classification is that of Lancefield, who discovered antibodies produced to a somatic carbohydrate of streptococcal organisms. Streptococci can be divided into groups according to the particular carbohydrate they possess on the basis of organism reaction to the different Lancefield antibodies (antisera). These groups are given a capital letter name ranging from A to G (in some classifications, even letters further in the alphabet but excluding E). Lancefield grouping does not depend on the presence of hemolysis or the type of hemolysis; for example, streptococci of group A are all beta hemolytic but streptococci of group D can be either alpha or beta hemolytic or even gamma nonhemolytic. Lancefield grouping cannot be done on some streptococci that are not beta hemolytic with the exception of those from groups B and D. Lancefield group A organisms, as mentioned earlier, are always beta hemolytic, and colonies are definitively identified with group-specific antiserum. This requires culture of a specimen, isolation of several colonies of a beta hemolytic organism resembling Streptococcus by colony appearance or Gram stain, and testing a colony with the Lancefield antisera. However, Lancefield antisera are relatively expensive, and since group A organisms seem to have an unusually marked susceptibility to the antibiotic bacitracin, Lancefield grouping is very frequently replaced by a less expensive identification method based on demonstrating inhibition of growth around a disk impregnated with a standardized concentration of bacitracin in an agar culture containing a pure growth of the organism (the organism is first identified as a Streptococcus by Gram stain or biochemical tests). However, this method is only presumptive rather than definitive, since about 5% (range, 4%-10%) of hemolytic group A streptococci are not inhibited by bacitracin (these organisms would be incorrectly excluded from group A), whereas 8%-22% of beta hemolytic streptococci from groups other than A (e.g., B, C, D, and G) have been reported to be sensitive to bacitracin (and therefore would be incorrectly assigned to group A.) Also, the bacitracin method takes 2 and sometimes even 3 days—1 day to culture the organism, 1 day to perform the disk susceptibility test, and sometimes another day to separate the organism colony from other bacterial colonies before the bacitracin test if the original culture grows several different organisms.

Group A organisms may be further separated into subgroups (strains) by use of special antisera against surface antigens (M antigens). Strain typing is mostly useful in epidemiologic work, such as investigating outbreaks of acute glomerulonephritis, and is not helpful in most clinical situations. Group A streptococci produce certain enzymes, such as streptolysin-O, which can be detected by serologic tests (antistreptolysin-O titer, Chapter 23). Antibodies against these enzymes do not appear until 7-10 days after onset of infection, so they are usually not helpful in diagnosing acute streptococcal infection. Their major usefulness is in the diagnosis of acute rheumatic fever and acute glomerulonephritis.

Rapid immunologic identification tests from many manufacturers have now become commercially available for group A streptococci that directly test for the organisms in swab specimens without culture. Usually the organism is extracted from the swab chemically or enzymatically and then tested with antiserum against group A antigen. The rapid tests can be performed in about 10-15 minutes (range, 5-60 minutes, not counting setup time of perhaps 5-15 minutes). Compared to throat culture, average overall sensitivity of the new direct methods is about 87%-95% (range 61%-100%, including different results on the same manufacturer’s kits by different investigators). Reported sensitivity is about 5%-10% higher if comparison culture specimens contain at least 10 organism colonies than if the culture density is less than 10 colonies. There is debate in the literature whether throat cultures with density of less than 10 colonies represent clinically significant infection. Some consider growth less than 10 colonies to represent a carrier state, but others disagree. Another point to remember is that rapid test sensitivity quoted in reports is not true clinical sensitivity, since rapid test sensitivity is usually less than culture sensitivity and culture sensitivity itself is rarely more than 95%. (It is probably less, since various changes in usual laboratory culture technical methodology for throat cultures have each been reported to increase culture sensitivity about 5%-10%, sometimes even more. In fact, one study on tonsillectomy patients found only 82% of preoperative throat cultures were positive when operative tonsillar tissue culture obtained group A streptococci.)

Lancefield group A streptococci are also known as Streptococcus pyogenes. Certain strains are associated with specific diseases, such as acute glomerulonephritis and acute rheumatic fever. Group A beta streptococci also produce various infections without any stain specificity. The most common is acute pharyngitis. Wound infections and localized skin cellulitis are relatively frequent. Other diseases that are much less common, although famous historically, include scarlet fever, erysipelas (vesicular cellulitis), and puerperal fever. There are even a few reports describing patients with necrotizing fasciitis or a condition resembling staphylococcal toxic shock syndrome.

Group A beta hemolytic streptococci may be isolated from throat or nasopharyngeal cultures in 15%-20% (range, 11%-60%) of clinically normal children. Nevertheless, for clinical purposes they are not considered normal inhabitants of the throat or nasopharynx since the usual current culture or serologic methods, which are done in the acute stage of infection, cannot reliably differentiate between carrier state and true pathogen, and prompt therapy for group A streptococcal nasopharyngeal infection is thought to decrease the possibility of acute rheumatic fever or acute glomerulonephritis. The accuracy of group A streptococcal isolation from throat cultures can be enhanced in several ways. Anaerobic rather than aerobic incubation has been reported to increase sensitivity about 15% (range, 0%-35%). Obtaining specimens on two swabs instead of one is reported to increase yield approximately 10%. Use of special differential media (with certain antibiotics added) helps to suppress growth of other organisms (e.g., hemolytic Haemophilus species) that may simulate beta hemolytic streptococcal colonies. However, not all of these enhancement techniques produce the same reported additive effect if they are combined. Although group A streptococci are almost always sensitive to penicillin on in vitro antibiotic sensitivity tests, 10%-30% of adequately treated patients continue to harbor the organisms and are considered carriers.

Lancefield group B streptococci (also known as Streptococcus agalactiae) are one of the most common causes of neonatal septicemia and meningitis. Escherichiacoli is the most frequent etiology of neonatal meningitis, but it has been reported that about one third of cases are due to group B streptococci. Neonatal group B streptococcal infection may occur in two clinical forms: early onset (before age 7-10 days, usually within 48 hours of birth), and late onset (after age 7-10 days). Septic neonates may have respiratory distress that mimics the noninfectious respiratory distress syndrome. The source of Streptococcus infection is the maternal genital tract, with 20%-30% (literature range, 4.6%-40%) of mothers and a similar percentage of all women of child-bearing age being culture-positive. Vaginal and perirectal colonization are more likely than cervical or urinary tract colonization. Colonization exists in all three trimesters in approximately the same percentage of women. About two thirds of women colonized in the first trimester of pregnancy are still culture-positive at delivery.

Group B streptococcus (GBS) causes about 20% of postpartum endometritis and 10%-20% of postpartum bacteremia. Caesarian section in a colonized mother has a higher risk of endometrial infection. GBS maternal colonization can also result in maternal urinary tract infection or bacteremia.

About 50% of infants born to mothers with culture-proven GBS become colonized (range, 40%-73%), with the bacteria usually acquired during delivery. About 1%-2% of colonized infants develop clinical GBS infection. GBS infection accounts for 30%-50% of neonatal serious infections. Principal risk factors for neonatal clinical GBS infection are prematurity and premature rupture of the membranes (also a twin with infection, maternal urinary tract infection or bacteremia, multiple births, or African American race). About two thirds of infected (not only colonized) neonates develop early onset disease and about one third develop late onset disease.

Diagnosis is most often made by maternal culture. The highest yield is obtained by one or more swabs from multiple sites in the distal third of the vagina plus the perineum (perirectal) area placed into selective liquid culture medium (usually Todd-Hewitt broth). Use of cervical cultures and solid culture media results in a considerably lower percentage of positive cultures. Rapid immunologic tests are available, similar in methodology to those for Group A streptococcal throat tests. However, current kit sensitivity for vaginal-rectal GBS overall is only about 40%-60% (range, 20%-88%), with the highest rates occurring in more heavily colonized patients. There is not a dependable cor- relation between degree of colonization and infant infection rate, although heavy colonization is more likely to result in neonatal infection.

There is controversy whether all mothers should be screened for GBS colonization before delivery (usually at about 26 weeks’ gestation), at delivery only, at delivery only if risk factors are present, or no screening at all with all infants being watched closely for the first hours after birth. There is also controversy whether those who are culture-positive early in pregnancy should be treated immediately, whether those positive at delivery should be treated, or if no cultures are done, whether prophylactic antibiotics should be given to patients with risk factors or to all patients during delivery. In general, positive cultures obtained early in pregnancy have a 70%-80% chance of a positive culture at delivery, whereas a negative culture obtained early in pregnancy is much less reliable in predicting status at delivery. Intradelivery parenteral antibiotics (usually penicillin or ampicillin) beginning before delivery, are the most commonly recommended method of prophylaxis. The two most frequent recommendations for this prophylaxis are a positive culture before delivery or presence of maternal risk factors. One recent report advocates intradelivery treatment for all mothers as the most cost-effective procedure (based on review of the literature).

GBS may also produce adult infections in nonpregnant persons (although this is not frequent), with postpartum endometritis, infection after urinary tract or gynecologic operations, pneumonia, and soft tissue infections being the most common types. Many affected adults have some underlying disease or predisposing cause. GBS also produces mastitis in cows.

Culture is the major diagnostic test for GBS. Besides culture, rapid latex agglutination slide tests for group B streptococcal antigen in body fluids are now available from several manufacturers. Although not many studies are published, reports indicate that these tests when performed on concentrated urine specimens detect 90%-100% of cases in culture-positive neonatal group B streptococcal bacteremia. Serum or unconcentrated urine is generally less successful.

Lancefield group C streptococci are primarily animal pathogens. However, these organisms may be found in the nasopharynx (1.5%-11%), vagina, and gastrointestinal (GI) tract of clinically normal persons. The most common strain in humans is Streptococcus anginosus. Group C streptococci occasionally produce human disease, most commonly (but not limited to) pharyngitis and meningitis (about 6% of cases; range, 3%-26%).

Lancefield group D streptococci contain several species, of which Streptococcus faecalis (one of the enterococci) and Streptococcus bovis (one of the nonenterococci) are the most important. Group D streptococci are responsible for approximately 20% (range, 8%-21%) of infectious endocarditis, typically that subgroup formerly called subacute bacterial endocarditis (SBE), and about 10% of urinary tract infections, as well as constituting the third most common cause of biliary tract infections. Group D streptococci are also associated with mixed wound infections, intraabdominal infections and a wide variety of other conditions, although there is some controversy about their relative pathogenicity in these circumstances. The majority of serious group D infections are due to enterococci. Enterococci are certain species of group D streptococci that are found normally in the human intestinal tract and that have certain special laboratory characteristics. These include resistance to heating and growth in certain media such as bile-esculin and 6.5% sodium chloride. Some species of enterococci produce alpha hemo- lysis, some produce beta hemolysis, and some are (gamma) nonhemolytic. Besides their role in infectious endocarditis (10%-12% of cases, range, 6%-20%), enterococci assume importance because they frequently are involved in other nosocomial (hospital-acquired) infections, particularly urinary tract infections (about 10%, range 6%-16%) and because they are usually partially resistant or resistant to penicillin. Bacteremia due to nonenterococcal S. bovis has been very frequently associated with either benign or malignant colon neoplasms (about two thirds of cases, range 17%-95%). Recently it has been proposed that the enterococci should be removed from the streptococci and placed into a new genus called Enterococcus. The new genus would include Enterococcus faecalis, Enterococcus faecium, Enterococcus durans, and several others. Streptococcus group D would include S. bovis and Streptococcus equinus.

Lancefield group F streptococci (Streptococcus anginosus also called milleri) form tiny colonies. Infection is uncommon, but they have been reported to cause bacteremia, dental abscesses, and abscesses in other body tissues, usually superimposed on preexisting abnormality.

Lancefield group G streptococci are among the normal flora in the nasopharynx (up to 23% of persons), skin, vagina, and GI tract. The most common serious infection is bacteremia, being isolated in 3.5% of all patients with bacteremia in one study. Many of these patients had serious underlying disease, especially cancer (21%-65% of patients in this study with group G bacteremia). Other infections are uncommon, but pharyngitis and arthritis are among the more important. Other Lancefield groups have occasionally been reported to produce human infection.

Viridans streptococci. The viridans group has been defined in several ways. The most common definition is streptococci that lack Lancefield antigens and produce alpha hemolysis on sheep blood agar. However, other investigators state that some viridans organisms are nonhemolytic, and still others say that some species can react with various Lancefield group antisera, although not with group B or D. Viridans streptococci are the most common cause of infectious endocarditis (that subgroup formerly known as SBE), being isolated in approximately 35%-40% of cases (range 30%-52%). They may also cause urinary tract infection, pneumonia, and wound infection. They are normal inhabitants of the mouth and may have a role in producing dental caries.

Streptococcus pneumoniae. This organism is clinically known as pneumococcus and occasionally is called Diplococcus pneumoniae in the literature. Although S. pneumoniae is a member of the Streptococcus genus, most nonmicrobiologists usually think of pneumonococcus as a separate entity from the other streptococci. Pneumococci are gram-positive diplococci that can be found in the throat of 30%-70% of apparently normal persons. They are still the most common cause of bacterial pneumonia, usually comprising at least 50% (range, 26%-78%) of community-acquired cases. They also produce many cases of middle ear infection in children and are an important cause of meningitis in older children, adolescents, and adults, most commonly in debilitated persons. Pneumococcal bacteremia develops in about 20% (range, 15%-25%) of patients with pneumococcal pneumonia, and pneumococcus is found in about 18% of all bacteremias, with the incidence being even higher in children. Splenectomy (or sickle cell anemia hyposplenism) is one well-known predisposing cause. The great majority of pneumococcal strains are very sensitive to penicillin, although about 4%-5% (range, 0%-20%, depending on the geographical area) of isolates are now relatively resistant and 1%-2% (range, 0%-6%) are resistant. Relative resistance is important in blood and cavity fluid, and especially in cerebrospinal fluid (CSF).

Pneumococci usually produce alpha (green) incomplete hemolysis on blood agar and thus mimic viridans streptococci. Morphologic differentiation from streptococci may be difficult, especially in cultures, since streptococci may appear singly or in pairs (instead of chains), and pneumococci often do not have the typical lancet shape or grouping in pairs. In the laboratory, differentiation is readily made because of the special sensitivity of pneumococci to a compound known as optochin. A disk impregnated with optochin is placed on the culture plate; inhibition of an alpha hemolytic coccus denotes pneumococci.

Besides diagnosis by culture, rapid slide latex agglutination tests for pneumococcal antigen are now available (developed for testing of CSF in patients with meningitis). Unfortunately, in the few reports available, detection of antigen in serum or unconcentrated urine has been less than 40%. Sensitivity in concentrated urine may be considerably better, but too little data are available.

Staphylococci

Staphylococci are gram-positive cocci that typically occur in clusters. Originally they were divided into three groups, depending on colony characteristics on blood agar: Staphylococcus albus (Staphylococcus epidermidis), with white colonies; Staphylococcus aureus, with yellow colonies; and Staphylococcus citreus, with pale green colonies. In that classification, S. aureus was by far the most important; it was generally hemolytic, and the pathogenic species were coagulase positive. The newer classification recognizes that coagulase activity (the ability to coagulate plasma) is a better indication of pathogenicity than colony color, since a significant number of coagulase-positive organisms are not yellow on blood agar. Therefore, all coagulase-positive staphylococci are now called S. aureus. Many microbiology laboratories still issue reports such as “S. aureus coagulase positive”; it is not necessary to include the coagulase result because S. aureus by definition is coagulase positive. Detection of heat-stable anti-DNA (“thermonuclease”), another enzyme produced by S. aureus, is generally considered to be the best confirmatory test for S. aureus should this be necessary. This test is nearly as sensitive and specific as coagulase but is more expensive and time consuming.

At this point, a few words should be said regarding coagulase tests. S. aureus produces two types of coagulase, known as bound coagulase (“clumping factor”) and free coagulase. The standard procedure is known as the tube coagulase test, based on the ability of free coagulase to clot plasma. This test requires 4 hours’ incubation and sometimes as long as 24 hours to confirm nonreactive results or weak positive reactions. Bound coagulase alone was originally used in a slide coagulase test that required only 1-2 minutes but produced about 20% false negative results. More recently, rapid 15- to 60-second slide tests based on reagents that react with either bound coagulase, protein A (another substance produced by S. aureus), or both have been introduced. Evaluations have shown sensitivities of about 98%-99% (range, 94%-100%) for these tests. However, false positive results have been reported in a few percent of most such tests, although the majority of evaluations of each test have not reported false positive results. Also, at least one investigator has reported at least 1% (and sometimes more) false negative results with each test when used with methicillin-resistant S. aureus. The family Micrococcaceae contains several genera, including Staphylococcus and Micrococcus. Both of these genera are composed of organisms that are identical on Gram stain and have certain biochemical similarities. As noted above, those that are coagulase positive are placed in the Staphylococcus genus and designated S. aureus. Gram-positive cocci that resemble S. aureus morphologically but that are coagulase negative have not yet been subjected to a uniform method of reporting. Some laboratories call all of them “S. epidermidis” or “coagulaseegative staphylococci.” Others differentiate organisms from the genus Staphylococcus and the genus Micrococcus on the basis of certain biochemical tests such as the ability of staphylococci to produce acid from glucose anaerobically. In general, non-aureus staphylococci, although much less pathogenic than S. aureus, are more pathogenic than micrococci. Some of the larger laboratories differentiate species of non-aureus staphylococci using certain tests. The species most commonly associated with human disease are S. epidermidis and Staphylococcus saprophyticus.

Staphylococci, as well as the enteric gramegative rod organisms and some of the Clostridia gram-positive anaerobes (to be discussed later), are normal inhabitants of certain body areas. The habitat of staphylococci is the skin. Therefore, a diagnosis of staphylococcal infection should not be made solely on the basis of S. aureus isolation from an external wound; there should be evidence that S. aureus is actually causing clinical disease. Besides the skin, about one half of all adults out- side the hospital carry S. aureus in the nasopharynx; this reportedly increases to 70%-80% if cultures are performed repeatedly on the same population over a long period. More than 50% of hospitalized persons have positive nasopharyngeal cultures. Exactly what factors induce these commensal organisms to cause clinical disease is not completely understood.

Staphylococcus aureus. Staphylococcus aureus is typically associated with purulent inflammation and characteristically produces abscesses. The most common site of infection is the skin, most frequently confined to minor lesions, such as pustules or possibly small carbuncles, but occasionally producing widespread impetigo in children and infants. The most frequent type of serious staphylococcal disease (other than childhood impetigo) is wound infection or infection associated with hospital diagnostic or therapeutic procedures.

In a small but important number of cases, S. aureus produces certain specific infections. Staphylococcal pneumonia may occur, especially in debilitated persons, or following a viral pneumonia. Meningitis and septicemia are also occasionally found, again, more commonly in debilitated persons or those with decreased resistance—often without any apparent portal of entry. S. aureus produces about 20%-25% (range, 9%-33%) of infectious endocarditis (especially that subgroup that used to be called acute rather than subacute). S. aureus causes a type of food poisoning different from that of the usual infectious agent; symptoms result from ingestion of bacterial toxins rather than from actual enteric infection by living organisms.

S. aureus frequently produces an enzyme known as beta-lactamase which makes the organism resistant to certain antibiotics that contain a beta-lactam ring structure such as penicillin G and ampicillin. S. aureus resistant to methicillin (MRSA) is a particularly difficult problem and will be discussed later in the section on antibiotic sensitivity testing. The nationwide incidence of MRSA (as a percentage of all S. aureus isolates) is about 4%; for individual hospitals, the range is 0%-60%. Staphylococcus epidermidis in many areas of the country is frequently resistant to a considerable number of antibiotics, including methicillin.

Toxic shock syndrome. The toxic shock syndrome is a disease strongly linked to S. aureus infection. The syndrome has been seen in various age groups in both sexes but occurs predominantly in females of childbearing age. There is an association with use of vaginal tampons, but a considerable number of cases have occurred when tampons were not used. Many (but not all) cases occur during the menstrual period. At least 15% of cases are not related to menses or tampon use but in- stead are preceded by surgical wound infection or by infections of other types. Clinical criteria for the syndrome include temperature of 102°F (38.9°C) or greater, erythematous sunburnlike or macular rash with eventual skin peeling or desquamation (especially on the palms and soles), hypotension or shock, clinical or laboratory involvement of at least four organ systems, and no definite evidence of bacteremia.

Involvement of organ systems is evidenced by a combination of symptoms and laboratory data. The classic syndrome includes at least four of the following organ systems (the approximate incidence of laboratory abnormality reported in one study is listed after each test):

1. Gastrointestinal: Vomiting and watery diarrhea.
2. Central nervous system: Headache, confusion, and disorientation. Stiff neck may occur. CSF tests are usually normal.
3. Liver: Aspartate aminotransferase (SGOT, 75%) or alanine aminotransferase (SGPT, 50%) more than twice the upper limit of the reference range. Total bilirubin may also be increased (70%).
4. Kidney: Blood urea nitrogen (BUN, 68%) or serum creatinine (90%) levels more than twice the upper limit of the reference range. There may be increased numbers of urinary white blood cells (WBCs) (73%) with negative urine cultures. Oliguria may occur.
5. Mucous membranes: Conjunctivitis.
6. Hematologic: Hemoglobin level is usually normal (77%). The WBC count may be normal (52%) or increased (48%). Neutrophil percentage is usually increased (86%), and frequently there are increased band forms. Dohle bodies are often present. Schistocytes are present in some patients (48%). The platelet count is less than 100,000/mm3 in 42%, the activated partial thromboplastin time is elevated in 46%, and fibrin split products are frequently elevated; but the fibrinogen level is normal in 86% and the prothrombin time is normal in 93%.
7. Muscular: Myalgias; creatine phosphokinase level is more than twice the upper limit of the reference range (63%).

It is not clear how many of the laboratory abnormalities are due to the disease itself or are secondary to hypotension with tissue hypoxia.

Staphylococcus aureus has been cultured from the cervix or vagina in approximately 75% of toxic shock patients (literature range 67%-92%) as opposed to approximately 7% (literature range, 0%-15%) of normal persons. Some series include culture data from the nasopharynx in addition to the cervix-vaginal region. This is more difficult to interpret since about 20% of normal persons carry S. aureus in their nasopharynx. Blood cultures are usually sterile. Toxic shock S. aureus is said to be usually nonhemolytic phage type 1, able to produce toxic shock toxin-1 (TSST-1) exotoxin, sensitive to erythromycin but resistant to penicillin. Currently, there is no readily available test to differentiate toxic shock from other entities with similar clinical or laboratory manifestations. It is possible to culture the organism and send an isolate to some reference laboratory able to test for production of TSST-1, but this would take several days.

Staphylococcus epidermidis. Staphylococcus epidermidis is found in more abundance on the skin than S. aureus and is a relatively frequent contaminant in cultures from the skin area. It may also be a contaminant in blood cultures or other cultures when the material for culture is obtained by needle puncture of the skin. However, S. epidermidis may produce disease. Infections most frequently associated with S. epidermidis are related to indwelling catheters, vascular grafts, or joint prosthetic devices; bacteremia in patients who are immunosuppressed; and infection after eye surgery (endophthalmitis). It is the causal agent of infectious endocarditis in 2%-3% of all endocarditis cases (range, 1%-8%), more often in persons with an artificial heart valve (where it may comprise up to 40% of prosthetic valve endocarditis). S. epidermidis is the most frequent cause of bacterial colonization or infection associated with indwelling vascula catheters and those used for hyperalimentation, where it is reported to cause about 20% of bacteremias and 40% of septicemias. Other catheter-related problem areas include CSF shunts (causing about 50% of infections; range, 44%-70%) and peritoneal dialysis catheters (about 50% of infections). S. epidermidis is also reported to cause about 40% (range, 25%-70%) of orthopedic joint prosthesis infections. Since S. epidermidis is a common skin-dwelling organism, it is a frequent culture contaminant. When it is obtained from a blood culture, there is always a question regarding its significance. Some rules of thumb from the literature that would tend to suggest contamination are isolation from only one of a series of cultures, isolation from only one of two cultures drawn close together in time, and isolation from only one bottle of an aerobic and anaerobic two-bottle culture set inoculated from the same syringe. However, none of these conclusively proves that the organism is a contaminant. Whereas S. aureus is differentiated from other staphylococci by a positive coagulase test result, many laboratories do not perform tests to identify the exact species of coagulaseegative staphylococci but report them all as S. epidermidis. However, not all of these organisms are S. epidermidis. Also, at least one of the coagulaseegative nonepidermidis staphylococci, S. saprophyticus, has been implicated as the causal agent in about 20% of urinary tract infections occurring in young women. On the other hand, S. epidermidis is not considered to be a urinary tract pathogen in young women, at least by some investigators. Therefore, it might be better to use the term “coagulaseegative Staphylococcus” in culture reports if the laboratory does not speciate these organisms. S. epidermidis is often resistant to many antibiotics, with 33%-66% reported to be resistant to methicillin.