INTRODUCTION — Escherichia coli that contain genes encoding Shiga toxins 1 and/or 2 are important human pathogens. Some of them, such as E. coli O157:H7, cause severe disease, most notably painful bloody diarrhea and the hemolytic uremic syndrome (HUS). These Shiga toxin-producing E. coli (STEC) infect people of all ages, but children under 10 years old and older adults have the most severe courses. Optimal management entails provider awareness at the point of presentation, rapid and accurate diagnostic microbiology, intravenous volume expansion, judicious and strategic laboratory testing, avoidance of potentially harmful interventions (eg, antibiotics and agents that impair gut motility or renal blood flow), careful monitoring, and engagement of public health disease control authorities.
Most clinical data on STEC infections are derived from studies in children. Most of the recommendations in this topic are thus informed by these data and our personal experience managing infection in children. However, STEC infections, HUS, and even fatal HUS can affect people of all ages. Clinical experience and the limited data on adult patients suggest a largely similar course of STEC infection among children and adults, with only few substantial differences [1]. Thus, our recommendations refer to both children and adults unless otherwise specified.
The clinical manifestations, diagnosis, and management of STEC infections are discussed here. The microbiology and epidemiology of STEC infections are discussed in detail elsewhere. (See "Shiga toxin-producing Escherichia coli: Microbiology, pathogenesis, epidemiology, and prevention".)
The clinical approach to STEC-associated HUS is also discussed in detail elsewhere. (See "Clinical manifestations and diagnosis of Shiga toxin-producing Escherichia coli (STEC) hemolytic uremic syndrome (HUS) in children" and "Treatment and prognosis of Shiga toxin-producing Escherichia coli (STEC) hemolytic uremic syndrome (HUS) in children".)
TERMINOLOGY — STEC fall into two clinically relevant categories: those that contain a gene encoding Shiga toxin 2 (with or without a gene encoding Shiga toxin 1) and those that do not (ie, their only Shiga toxin gene encodes Shiga toxin 1) (table 1). STEC containing Shiga toxin 2 and its allelic variants are more virulent than those containing only Shiga toxin 1 and/or its variants. STEC that contain the gene encoding Shiga toxin 2 are often associated with bloody diarrhea and can cause hemolytic uremic syndrome (HUS) independent of serogroup, whereas those that only contain the gene encoding Shiga toxin 1 are very rarely associated with bloody diarrhea and carry minimal or no risk for HUS. (See "Shiga toxin-producing Escherichia coli: Microbiology, pathogenesis, epidemiology, and prevention", section on 'Shiga toxin type'.)
Almost all E. coli O157:H7 contain a gene encoding Shiga toxin 2 (and about two-thirds additionally contain a gene encoding Shiga toxin 1) [2,3]. This serotype remains the leading cause of STEC infections worldwide and is the serotype most likely associated with epidemics and severe disease. Many STEC that belong to other serotypes (ie, non-O157:H7 STEC) contain a gene encoding Shiga toxin 1 but not Shiga toxin 2; only a subset of non-O157:H7 serotypes isolated from humans can produce Shiga toxin 2 and have the capacity to cause severe disease. Paradoxically, among the STEC that contain a gene encoding Shiga toxin 2, those that also contain a gene encoding Shiga toxin 1 are somewhat less virulent (in terms of HUS risk) [4,5].
In this topic, the term "high-risk STEC" refers to E. coli O157:H7 and the subset of non-O157:H7 serotypes that contain a gene encoding Shiga toxin 2. This includes sorbitol-fermenting E. coli O157:NM, a pathogen largely confined to Germany and the Czech Republic. Although most clinical literature pertains to infections caused by E. coli O157:H7, the data can be confidently extrapolated to apply to infections caused by any non-O157:H7 STEC that contains a gene encoding Shiga toxin 2.
The terminology of STEC, which can be confusing and problematic, is discussed in detail elsewhere. (See "Shiga toxin-producing Escherichia coli: Microbiology, pathogenesis, epidemiology, and prevention", section on 'Microbiology'.)
CLINICAL FEATURES
Incubation period — Following exposure, STEC infections have a median incubation period of three days, with a range of 1 to 10 days. This estimate is based on E. coli O157:H7 outbreaks in which the time and date of the exposure can be determined [6]. However, many STEC infections cannot be linked to a specific vehicle or exposure, so the exact timing of ingestion is often unknown.
Typical clinical findings — Infections with high-risk STEC (ie, E. coli O157:H7 or non-O157 STEC that encode a gene for Shiga toxin 2) are typically characterized by painful diarrhea that becomes bloody, often without persistent fever. (See 'Terminology' above.)
The course of infection with high-risk STEC follows a patterned sequence of events (figure 1) [7]. We consider the first day of illness to be the first day of diarrhea (and we base management decisions on the day of illness (algorithm 1)), although abdominal pain, vomiting, and fever can precede the onset of diarrhea. The initially non-bloody diarrhea typically becomes visibly bloody one to three days after onset; however, approximately 15 to 20 percent of patients with E. coli O157:H7 experience only non-bloody diarrhea (data are sparser for high-risk non-O157 STEC). Uncommonly, patients with E. coli O157:H7 report only a few hours of non-bloody diarrhea before the stool becomes visibly bloody or, even more rarely, state that the first loose stool was bloody [8]. Diarrhea generally resolves after about seven days.
Hemolytic uremic syndrome (HUS), if it occurs, is usually established 5 to 13 days after the onset of diarrhea. In very rare cases, patients infected with high-risk STEC have no diarrhea but still develop HUS [9-11]. (See 'Hemolytic uremic syndrome' below.)
It is very unusual for patients with STEC infection to be febrile at presentation, but many patients or their families or caregivers report a fever early in the illness, prior to seeking care. As an example, in a retrospective review of 260 cases of STEC infection, fever was self-reported in approximately 30 and 16 percent of O157:H7 and non-O157 E. coli cases, respectively [12], but despite such reports of pre-presentation fever, documented fever in medical settings is quite uncommon [13]. Absence of fever at presentation does not distinguish infections caused by E. coli O157:H7 from other bacterial enteric infections [14]. Nevertheless, high fever (>38.5°C) in a patient with painful diarrhea makes other pathogens (eg, Shigella, Campylobacter, or Salmonella) the more likely etiologic agents. (See 'Clinical suspicion' below.)
Other highly prevalent features of E. coli O157:H7 infection include abdominal pain and an acute onset [8,15]. However, these do not distinguish high-risk STEC infections from bloody diarrhea caused by other pathogens [14].
When bloody diarrhea first develops during a high-risk STEC infection, the platelet count, creatinine concentration, and hematocrit are typically normal, and there is no evidence of hemolysis on peripheral blood smear. However, abnormalities in coagulation and fibrinolysis suggestive of vascular injury can be observed at this early point in illness and, as infection progresses, the platelet count often falls, sometimes to values below the lower limit of normal, even if the patient does not develop HUS [16-20]. We use the day-to-day trend in platelet counts as an indicator of risk of HUS. (See 'Timing of fluid discontinuation and discharge' below.)
Complications
Hemolytic uremic syndrome — The most consequential complication of STEC infection is hemolytic uremic syndrome (HUS). HUS is characterized and diagnosed by the triad of nonimmune-mediated hemolytic anemia, thrombocytopenia, and acute kidney injury (we recommend thresholds of hematocrit <30 percent, platelet count <150,000 microL, and serum creatinine greater than the upper limit of normal for age).
The likelihood that a patient infected with E. coli O157:H7 develops HUS varies by age. Fifteen to 20 percent of children younger than 10 years with culture-proven E. coli O157:H7 infection develop HUS. The HUS rate after childhood is difficult to determine with certainty because ascertainment of infection rates might differ between children and adults, and STEC-associated HUS may be misclassified in adults as atypical HUS or thrombotic thrombocytopenic purpura. Nevertheless, although the incidence of diagnosed STEC infection falls markedly during and after adolescence, STEC infections that occur in adults are not categorically more benign than in children. As an example, in a 2018 North American outbreak in which the majority of the 210 infected cases were adults, 96 were hospitalized, 27 developed HUS, and 5 died [21].
Patients who develop HUS generally meet the case definition on a median of day 7 to 7.5 of illness, with a range between 5 and 13 days (the first day of illness is the first day of diarrhea (figure 1)) [22,23]. We have occasionally treated patients infected with E. coli O157:H7 who develop HUS on day 4 after the onset of diarrhea and are aware of one patient who developed HUS after several weeks of colonization with an E. coli O157:H7 that initially caused bloody diarrhea [24]. Rarely, patients with HUS caused by a high-risk STEC E. coli have no diarrhea [9-11].
The clinical manifestations and diagnosis of STEC-associated HUS are discussed in detail elsewhere. (See "Clinical manifestations and diagnosis of Shiga toxin-producing Escherichia coli (STEC) hemolytic uremic syndrome (HUS) in children", section on 'Clinical and laboratory manifestations' and "Clinical manifestations and diagnosis of Shiga toxin-producing Escherichia coli (STEC) hemolytic uremic syndrome (HUS) in children", section on 'Diagnosis'.)
Other — Very rarely, children and teenagers with an STEC infection present with intussusception [25]. In some institutions, antibiotics are administered before nonoperative reduction of intussusception, but in cases of intussusception associated with STEC, we discourage administering antibiotics. (See 'Antibiotics' below.)
Intussusception is discussed in detail elsewhere. (See "Intussusception in children".)
EVALUATION AND DIAGNOSIS
Clinical suspicion — Infections with high-risk STEC (ie, E. coli O157:H7 or non-O157 STEC that encode a gene for Shiga toxin 2) should be suspected in the following [9-11]:
●All patients of all ages with acute bloody diarrhea.
●Individuals who develop diarrhea (even if non-bloody) after contact with a patient with known STEC infection.
●All patients with hemolytic uremic syndrome (HUS), even if they do not have diarrhea.
Because high-risk STEC cause severe illness, we presumptively manage all patients with clinical suspicion for STEC infection as having infection, pending microbiologic testing. However, some features decrease the likelihood of a high-risk STEC infection (algorithm 1). Such features are:
●Absence of abdominal pain
●Diarrhea that abates within several hours of presentation
●Total duration of non-bloody and bloody diarrhea greater than one week
●Findings that suggest a chronic process (eg, microcytic anemia, weight loss)
●Contact with a patient with non-STEC bacterial colitis
Fever (arbitrarily >38.5°C or >101.3°F) at the time of presentation is also atypical but should not be used to discount the possibility of STEC infection.
We strive to confirm or refute a high-risk STEC infection with microbiologic testing as early in illness as possible, because early diagnosis lends clarity to a sometimes confusing clinical situation and can reduce the number of patients admitted to hospital unnecessarily. Because there are challenges to meeting this goal, we use an admit all approach at our center, with the recognition that most children with acute bloody diarrhea do not have a high-risk STEC infection. The approach includes at least a six-hour observation period. If, in that interval, abdominal pain abates or greatly diminishes, or the diarrhea becomes non-bloody, decreases in volume, or stops entirely, accelerated discharge with follow-up as an outpatient may be appropriate. However, during that observation period, we commence management of presumptive high-risk STEC infection with aggressive intravenous volume expansion (figure 2). Stopping points include improvement within hours of admission and exclusion of a high-risk STEC infection by microbiologic testing (algorithm 1). Rapid nonculture diagnostics can detect STEC within several hours, thereby reducing admissions and over-treatment of individuals with bloody diarrhea from non-STEC infections. (See 'Microbiologic diagnosis' below.)
Our approach to management of suspected and documented STEC infection is discussed in detail below. (See 'Management' below.)
A high index of suspicion is important, because children with STEC infection often have symptoms severe enough to present to emergency departments but account for a small proportion of patients with acute diarrhea in that setting; thus they are often not recognized as having STEC and being at risk for HUS. In one study of children with STEC presenting to pediatric emergency departments in the United States and Canada, one in seven developed HUS within a median of three days of that visit, and nearly 30 percent of those who subsequently developed HUS were initially sent home [26].
Initial evaluation — The initial evaluation of patients with suspected STEC infection incudes determination of the stage of disease, basic laboratory testing, and microbiologic testing for STEC and other bacterial pathogens associated with bloody diarrhea on rectal swab or stool specimens:
●When asking about presenting symptoms, we determine the first day that the patient experienced loose stools. Most publications designate this day as the first day of illness (figure 1); although prodromal symptoms such as vomiting or abdominal pain often precede the first day of diarrhea, they are too variable and nonspecific to serve as points of reference for subsequent phases of illness [27]. We also ask about visible blood in the stools and abdominal pain. On physical examination, we focus on vital signs, heart and lung auscultation (in anticipation of aggressive volume expansion), and abdominal examination.
●We check the complete blood count, electrolytes, and serum creatinine at presentation and use these data as baselines for reference on subsequent days. We do not routinely check a urinalysis, transaminase levels, coagulation tests, or circulating inflammatory markers early in illness, as they may be misinterpreted and prompt misguided interventions [27,28]. As an example, an abnormal urinalysis might prompt treatment of a presumed urinary tract infection if the urinalysis is abnormal, despite the likelihood that an abnormal urinalysis in the setting of infectious diarrhea reflects contamination of urine with stool. Serum lactate dehydrogenase (LDH) has been used to monitor progression and resolution of established thrombotic microangiopathies of various etiologies and could be a harbinger of hemolysis. While LDH might be useful in predicting outcome [26], data are not sufficient to start, stop, or modify fluid treatment based on LDH levels.
●Rapid and accurate microbiologic testing is critical for the management of suspected STEC infections and is an essential component of the initial evaluation. Specimen collection and microbiologic testing for STEC are outlined below. (See 'Microbiologic diagnosis' below.)
Computed tomography (CT) rarely, if ever, has a role in evaluating suspected or documented STEC infections in children; the low likelihood that the findings could inform management is outweighed by the risks of ionizing radiation and the potential nephrotoxicity of intravenous contrast. In adults, however, in whom noninfectious intestinal disorders can present with bloody diarrhea, CT (and other testing) might have value [29]. Nevertheless, in all cases of acute bloody diarrhea in adults, prompt and thorough microbiologic assessment is of paramount importance.
The initial evaluation of patients with acute diarrhea in general is discussed in detail elsewhere. (See "Diagnostic approach to diarrhea in children in resource-rich countries", section on 'Acute diarrhea (duration <5 days)' and "Approach to the adult with acute diarrhea in resource-rich settings", section on 'Evaluation'.)
Microbiologic diagnosis
Specimen collection — When STEC is suspected, a rectal swab or stool should be collected and submitted for STEC testing as a matter of priority. Unless the patient can immediately produce a stool specimen, we suggest using rectal swab specimens because of the importance of an expedited diagnosis.
Waiting for a stool delays submission of the specimen for testing. Furthermore, in the rare patient with HUS but no diarrhea, a rectal swab culture might be the only way to test for STEC. In a study of children with diarrhea, rectal swab specimens were associated with higher pathogen yields than stool specimens [30]. The Infectious Diseases Society of America (IDSA) recognizes the value of swab cultures of stool in children but does not endorse such a strategy for adults because of concerns about sensitivity [31]. However, we encourage rectal swab specimens in adults with bloody diarrhea if a bulk stool cannot be obtained, as a rectal swab culture can accelerate the time to microbiologic diagnosis [32].
Detecting STEC — The rectal swab or stool specimen should be submitted for selective culture for E. coli O157:H7 and Shiga toxin testing, either by immunoassay following overnight broth culture or by molecular technology [33].
The diagnosis of STEC infection is made in a patient with acute diarrhea when E. coli O157:H7 is isolated or when Shiga toxin (or a gene encoding a Shiga toxin) is detected. If a Shiga toxin or Shiga toxin gene is detected but E. coli O157:H7 is not identified, it is important to try to determine if the Shiga toxin is Shiga toxin 2, as this indicates a high-risk STEC and has prognostic and management implications. Until the toxin genotype is known, we take a cautious approach and consider an unspecified Shiga toxin signal from a patient with acute diarrhea to reflect a high-risk STEC pathogen (figure 2). (See 'Terminology' above.)
Available testing may vary by clinical laboratory, and each has advantages and drawbacks (table 2):
●Selective culture – All patients with suspected STEC infection should have specimens submitted for selective culture for STEC. Sorbitol-MacConkey (SMAC) agar is the prototypical selective medium used to isolate E. coli O157:H7. This serotype ferments sorbitol slowly and appears as translucent colonies on this medium (picture 1). Such colonies can be confirmed as E. coli biochemically and tested for reaction with antisera to the O157 antigen. Strains presumptively identified as E. coli O157:H7 should be sent to a reference laboratory for confirmation.
CHROMagar STEC is a different selective media that can identify E. coli O157:H7 as well as some non-O157:H7 STEC isolates, but it may not be as sensitive for certain non-O157:H7 STEC as antigen immunoassay following broth culture.
●Shiga toxin enzyme immunoassay (EIA) following incubation in broth culture – EIAs can detect E. coli O157:H7 as well as non-O157:H7 STEC but fail to detect approximately 10 to 15 percent of E. coli O157:H7 that are detected on agar plating [34-40]. Thus, Shiga toxin EIA testing can be used in addition to, but never in lieu of, selective culture. Newer EIA tests can differentiate between Shiga toxins 1 and 2.
●Molecular testing – Multiplex molecular panels that include nucleic acid amplification tests (NAATs) that detect the genes for Shiga toxins 1 and 2 are increasingly available and can deliver results within several hours [41]. Molecular detection of a Shiga toxin gene on a specimen from a patient with painful or bloody diarrhea or on a broth culture of the stool can be rapidly diagnostic of an STEC infection, with some caveats:
•Specimens should still be submitted for selective culture if a Shiga toxin gene is identified by NAAT. Pathogen isolation is critically important for epidemiologic analysis and outbreak tracing [42,43].
•Although many molecular tests can distinguish between genes encoding Shiga toxins 1 and 2, these details are not necessarily reported. If the molecular test result states that STEC is detected, details on toxin type should be requested from the microbiology laboratory. However, the ability to provide this information to the clinician might vary according to the test platform used and laboratory policy.
•In settings other than acute diarrhea (eg, persistent or chronic diarrhea), the clinical relevance of a positive STEC NAAT is uncertain and usually does not warrant specific clinical action.
•If NAAT (or other diagnostics) detects another enteric pathogen (eg, C. difficile, Campylobacter, Salmonella, or Shigella) in addition to a high-risk STEC, we defer antibiotics because of the risk of HUS (see 'Antibiotics' below). Multiple enteric infections are more likely in patients with non-O157:H7 STEC infections than with E. coli O157:H7 infections. However, some non-O157:H7, such as E. coli O111, can possess a gene encoding Shiga toxin 2, which again emphasizes the role of Shiga toxin genotyping in clinical decision-making.
Providers are highly reliant on accurate microbiology to manage patients with STEC infections, because the history, physical examination, and common laboratory tests cannot distinguish patients infected with E. coli O157:H7 from those infected with other bacterial enteric pathogens [14]. In our experience, the availability of reliable diagnostic microbiology results early in illness is strongly associated with favorable clinical outcomes in E. coli O157:H7 infections [44].
We thus encourage microbiology laboratories to set up cultures throughout the day and night, every day. This practice has been in place in our institution for over a decade and results in return of STEC testing results in 24 hours [14], a considerably shorter interval than reported in centers that have not yet adopted such policies [22]. We also encourage microbiology laboratories to alert clinicians to presumptive positive testing for STEC while awaiting confirmatory testing.
The spectrum of data that might be provided to the physician during the course of an STEC infection and our recommendations on management when this information becomes available are illustrated in the figure (figure 2).
Caveats to other microbiologic testing — Stool or rectal swab specimens are also evaluated for other bacterial pathogens that cause bloody diarrhea. In resource-rich settings, we discourage testing for pathogens that do not cause bloody diarrhea, in particular viruses and parasites.
If a multiplex panel was performed and identifies a nonbacterial pathogen in the setting of acute bloody diarrhea, we generally do not act on this result. An exception is detection of Entamoeba histolytica in an at-risk patient; however, in North America, such infections are quite rare in childhood.
We do not recommend testing for Clostridioides difficile, because this infection rarely presents as suddenly as STEC infection. Also, fulminant C. difficile colitis is particularly unusual in previously healthy children, and many children and adults excrete C. difficile without symptoms [45]. Depending on the laboratory, C. difficile results are returned before other bacterial pathogens have been excluded, and, if positive, might prompt use of an antibiotic (eg, metronidazole) that is associated with an increased risk for HUS [24,46,47]. (See 'Antibiotics' below.)
DIFFERENTIAL DIAGNOSIS — The differential diagnosis of acute bloody diarrhea predominantly includes other bacterial pathogens, which are discussed in detail elsewhere. (See "Causes of acute infectious diarrhea and other foodborne illnesses in resource-rich settings", section on 'Inflammatory diarrhea'.)
Assessing these and other causes of acute diarrhea in children and adults is presented separately. (See "Diagnostic approach to diarrhea in children in resource-rich countries", section on 'Acute diarrhea (duration <5 days)' and "Approach to the adult with acute diarrhea in resource-rich settings", section on 'Evaluation'.)
MANAGEMENT
Objectives — There are no specific treatments for STEC infections. Therefore, our approach to managing high-risk STEC (ie, E. coli O157:H7 and non-O157 STEC that contain a gene encoding Shiga toxin 2) infections focuses on practices that lower the likelihood of hemolytic uremic syndrome (HUS), particularly anuric HUS. Because of the serious nature of HUS, we apply this approach to patients with suspected, as well as confirmed, infections caused by high-risk STEC (algorithm 1). (See 'Clinical suspicion' above.)
We encourage the same approach to management in both children and adults. However, this approach may be less valuable for adults because, compared with children, adults are less likely to be infected with STEC and the rate of HUS in culture-confirmed, high-risk STEC infections is lower [48]. However, we are concerned that STEC infections in adults might have higher rates of HUS than previously appreciated, as demonstrated in one North American outbreak, in which the HUS rate exceeded 10 percent in a mostly adult cohort [21].
Timely and accurate microbiologic testing is also essential for appropriate management, and clear ongoing communication with the microbiology laboratory is a key component of care, although the clinician should be aware of potential difficulties with determining Shiga toxin genotype in a timely manner. (See 'Microbiologic diagnosis' above.)
Site of care — Because of the risk of HUS, we favor hospital admission for patients with suspected or confirmed infection with a high-risk STEC. Hospitalization permits aggressive intravenous fluid management and daily reassessment, which are the fundamental components of our management approach, as detailed in the sections that follow. Hospitalizing infected patients can also lower the rate of secondary infection in the community [49].
STEC infections are overall rare compared with the higher rate of gastroenteritis caused by other pathogens, and their management differs considerably. Thus, if possible, we encourage admission of patients with confirmed or suspected STEC infection to specific nursing units and care teams to maximize experience and offer optimized center expertise. Partnership with the diagnostic microbiologic laboratory is also a major element of caring for these patients.
Fluid management
Aggressive IV volume expansion — For patients with suspected or confirmed high-risk STEC infection, we suggest early and aggressive intravenous (IV) volume expansion with isotonic crystalloid. A peripheral IV catheter usually suffices, but in patients in whom peripheral edema develops, a peripherally inserted central catheter can forestall challenges with access. We ty to avoid antecubital phlebotomies to keep this option open.
We do not rely on oral fluids to maintain euvolemia; enteral electrolyte-containing solutions contain insufficient sodium to confer nephroprotection [44], and vomiting, which is common among patients with STEC infection who subsequently develop HUS complicates oral rehydration [47].
Evidence from observational cohort studies indicates a plausible benefit to early IV volume expansion among individuals with high-risk STEC infection [26,44,50-52]. In a meta-analysis of observational studies of children with STEC- or diarrhea-associated HUS, apparent volume depletion was associated with mortality, and IV fluid administration was associated with a decreased risk of renal replacement therapy (odds ratio [OR] 0.26, 95% CI 0.11-0.60) [52]. In two prospective cohort studies of children with HUS, pre-HUS volume expansion was associated with lower rates of anuria [44,50]. IV volume expansion has been implemented at initial presentation (ie, before HUS ensues) [53,54] and given to children with established HUS [55,56].
It is uncertain whether IV volume expansion early in illness is associated with reduced frequency of HUS following STEC infections or just reduced anuria, which almost always occurs by day 10 of illness [50]. Nevertheless, anuric HUS is categorically worse than non-anuric HUS, so averting anuria has merit even if HUS does develop [57-69]. In one study of children with STEC infection who were evaluated in pediatric emergency departments in North America, those who developed HUS (which occurred at a median of three days after presentation) had higher hematocrits and lower serum sodium concentrations at initial presentation than those who did not [26]. This finding reinforces the value of early recognition of infection and volume expansion, because relative hemoconcentration and hyponatremia are potentially manageable risk factors.
While observational data suggest that volume expansion early in illness may be associated with better outcomes if HUS occurs, a randomized controlled trial of IV volume expansion would provide the most convincing evidence of the efficacy and safety of this strategy. In the interim, we have used inpatient IV volume expansion to manage confirmed or suspected high-risk STEC infections.
Fluid type and content — We use isotonic crystalloid (ie, normal saline or a balanced salt solution such as Lactated Ringer solution) rather than hypotonic fluids because the goal of parenteral volume expansion is to preserve renal blood flow, which is better achieved with isotonic fluids. Hyponatremia at the time of HUS diagnosis is associated with worse outcomes, additionally justifying isotonic volume expansion [51]. Except for boluses, we use fluids containing 5% dextrose.
Vigorous volume expansion using crystalloid in which chloride is the exclusive anion (ie, normal saline) can contribute to iatrogenic hyperchloremic metabolic acidosis. We therefore prefer balanced salt solutions for volume expansion. However, balanced salt solutions contain sodium at concentrations that are considerably less than the 154 mEq/L of normal saline, so if the serum sodium concentration falls below normal, we revert to normal saline. Similarly, balanced salt solutions contain <5 mEq/L of potassium, considerably lower than what is generally added to normal saline (20 or more mEq/L). This can be remedied by either infusing potassium chloride from a separate bag and working with a pharmacist to increase the potassium concentration or using normal saline with added potassium.
If potassium is added to the fluid, we monitor serum electrolytes at least daily; even though STEC-associated HUS is characterized by brisk hemolysis and an imminent risk of renal insufficiency, we inexplicably observe hypokalemia more frequently than hyperkalemia.
Rate of administration — For patients with suspected or confirmed high-risk STEC infection, we initiate and maintain generous IV volume expansion with isotonic crystalloid. In children, we start with a bolus of at least 20 mL/kg of normal saline and administer 2 liters per m2 per day. Adults should also receive aggressive isotonic volume expansion, but the volume administered might be tempered by other clinical considerations, especially if patients have cardiopulmonary conditions that might be exacerbated by fluid administration.
Our goal for IV fluid administration is reduction of the hemoglobin concentration. To guide hemodilution, we obtain a complete blood count (CBC) every 8 to 12 hours early in illness to monitor the effectiveness of the volume expansion. (See 'Daily monitoring' below.)
The aggressiveness of our attempts to hemodilute patients depends on various factors, including the degree of hemoconcentration, day of illness, and effect of attempts to lower the hemoglobin. Examples of patients in whom we suggest especially vigorous volume expansion include:
●A patient with a hemoglobin of 17 g/dL and a normal platelet count on the third day of diarrhea. For such a patient, we aim for a hemoglobin concentration of 13 g/dL within 24 hours, and even lower 24 hours after that.
●A patient on day 4 of illness with a hemoglobin concentration of 12 g/dL that remains constant or increases over 24 hours despite aggressive volume expansion. Such a situation is concerning for early capillary leakage.
In both of these situations in children and adolescents, we use multiple 20 mL/kg boluses of isotonic crystalloid (generally normal saline) to increase or maintain perfusion to the kidneys, rather than increase the maintenance fluid infusion rate above 2 liters per m2 per day. Although we have insufficient experience to state that this approach is appropriate in adults, we believe that it should be applied because of the threat posed by renal hypoperfusion with high-risk STEC infections in patients of any age. Assiduous monitoring of cardiopulmonary status is important in adults, especially in people with pre-existing heart and lung conditions. Assessment of the pulse and respiratory rate and chest auscultation should be performed prior to fluid boluses in patients of all ages.
If there is any question of diminished urine output or lack of hemoglobin dilution, we also repeat boluses of normal saline (10 to 20 mL/kg), as long as the patient does not have signs of central volume overload. Peripheral edema should be expected and should not prompt fluid restriction in the absence of central volume overload.
The evolution from STEC infection into HUS follows a highly patterned sequence of events (figure 1), with a decrease in platelets preceding the drop in hemoglobin from hemolysis. Therefore, the day of illness and the trend in the platelet count can be used to determine if a decreased hemoglobin reflects adequate volume expansion versus early hemolysis. A falling hemoglobin concentration accompanied by a slightly falling or stable platelet count on or before day 4 of illness usually reflects desired hemodilution from volume expansion. However, a hemoglobin concentration of 12 g/dL on day 6 of illness in the setting of a platelet count <100,000/microL and normal serum creatinine concentration probably reflects relative hemoconcentration, since hemolysis has likely already started and the hemoglobin level would have been expected to be in the 8 to 10 g/dL range. In this latter situation, we attempt vigorous volume expansion but also carefully monitor the patient, as azotemia might ensue very quickly. Relatively elevated hemoglobin at the point in illness when HUS is diagnosed is a poor prognostic sign [51,55,57,70,71].
At all points in illness, volume contraction in patients infected with STEC cannot simply be attributed to fluid loss from diarrhea (STEC diarrhea is not voluminous). Capillary leakage can occur, often in advance of azotemia.
Fluid management in patients with HUS is discussed in detail elsewhere. (See "Treatment and prognosis of Shiga toxin-producing Escherichia coli (STEC) hemolytic uremic syndrome (HUS) in children", section on 'Fluid and electrolyte management'.)
Daily monitoring — Because infections with high-risk STEC evolve day by day and management is guided by the phase of illness, daily reassessment is critical [27]. Clinical assessment includes evaluation of symptoms (eg, whether the diarrhea is improved and abdominal pain has decreased), volume status, and urine output.
We perform a CBC every 8 to 12 hours during the first 24 hours of IV fluid administration to confirm that volume expansion is achieving its goal of therapeutic hemodilution (see 'Rate of administration' above); thereafter, we check this value at least daily. We also check electrolytes and creatinine daily, with more frequent checks if concerning abnormalities develop. Serum albumin concentrations are sometimes useful later in illness, especially if there is peripheral edema, as albumin infusions can be helpful to expand intravascular volume and reduce edema. (See 'Hypoalbuminemia' below.)
Daily monitoring is also important to identify HUS early, if this outcome develops. We strongly endorse a stringent, age-focused definition for HUS diagnosis that is based on the simultaneous occurrence of [46]:
●Hematocrit <30 percent
●Platelet count <150,000/microL
●Serum creatinine concentration greater than the upper limit of normal for the patient's age
Other UpToDate authors diagnose STEC-associated HUS more generally based on the triad of microangiopathic hemolytic anemia with fragmented red blood cells, thrombocytopenia, and acute kidney injury (as defined by a decrease in the estimated glomerular filtration rate). (See "Clinical manifestations and diagnosis of Shiga toxin-producing Escherichia coli (STEC) hemolytic uremic syndrome (HUS) in children", section on 'Diagnosis'.)
Timing of fluid discontinuation and discharge — For patients with confirmed infection caused by a high-risk STEC, discontinuation of IV fluids depends chiefly on the stage of the illness and the platelet count as well as clinical improvement (algorithm 1) (see 'Daily monitoring' above). We consider the first day of illness to be the first day of diarrhea (not the first day of bloody diarrhea) (figure 1):
●Days 1 through 4 of illness (early phase) – We generally continue IV fluids through the first four days of illness, as we do not have enough experience with trends in the clinical trajectory during this time to confidently halt volume expansion.
●Days 5 through 9 of illness (middle phase) – We discontinue IV fluids and discharge the patient if the platelets have increased by at least 5 percent compared with the day prior and the patient's clinical condition has improved (eg, diarrhea and abdominal pain have considerably improved). We check a CBC the following day.
●Day 10 or later of illness (late phase) – We discontinue IV fluids and discharge the patient if the platelets have not decreased by over 5 percent compared with the day prior and the patient's clinical condition has improved. We check a CBC the following day.
A decrease in platelets is often the first laboratory manifestation of HUS. Based on our clinical experience, rising platelet levels in the middle or late phases of illness suggest that the risk for HUS has passed. We have used the 5 percent threshold to manage many patients without adverse event, but it has not been systematically evaluated.
The post-discharge CBC is performed to ensure there is no decrease in the platelet count reflecting late deterioration. This is unusual, but if it does occur, we typically readmit patients for an additional day or two of IV fluids. We also obtain a repeat rectal swab or stool culture prior to discharge in patients who may need to document a negative culture prior to returning to daycare, school, or work. (See 'Return to school or work' below.)
For patients who were managed with volume expansion because of suspected STEC but had negative microbiologic testing and are clinically improving, IV fluid can be discontinued and the patient can be discharged, as long as discharge is otherwise appropriate.
Avoid potentially harmful interventions
Antibiotics — We recommend against treating confirmed or suspected STEC infections (either E. coli O157:H7 or non-O157:H7) with antibiotics. If an STEC is identified in patients in whom antibiotic therapy was initiated empirically, we discontinue it.
This recommendation is based on the association between antibiotics and the development of HUS in patients with STEC infection. The risk was illustrated by a meta-analysis of mainly observational studies that included over 1800 patients with STEC infection [72]. When all studies were pooled, there was a nonsignificant trend toward a higher risk of HUS with antibiotic use (pooled OR 1.33, 95% CI 0.89-1.99). However, among studies that were deemed to have low risk of bias and that employed a stringent definition of HUS, the association was stronger (pooled OR 2.24, 95% CI 1.45-3.36) [46]. One of the studies included was a prospective study of 259 children <10 years of age infected with E. coli O157:H7 [47]. HUS occurred more frequently in the 25 children who received antibiotics, which included trimethoprim-sulfamethoxazole, beta-lactams, metronidazole, and azithromycin (36 compared with 12 percent in those who did not receive antibiotics). After adjusting for other variables, the absolute antibiotic-attributable risk increase for HUS incidence was 25 percent, corresponding to one case of HUS for every four children treated with antibiotics. Data in adults are sparser; in three studies among adults that were included in the meta-analysis discussed above and were considered high quality, antibiotics administered early in STEC infection were associated with subsequent development of HUS [1,72-74].
In contrast to these data, we are unaware of any study that has convincingly demonstrated that antibiotics reduce the risk of developing HUS in people of any age with a high-risk STEC infection.
Antibiotics have not been shown to reduce symptoms or other complications associated with STEC infections [75].
Other medications — Apart from antibiotics, we avoid several additional medications that might be used in acute painful gastroenteritis because of lack of benefit and/or potential harm in patients with STEC infection:
●Antimotility agents (including opioids) – These agents have been associated with a higher risk of HUS and, among those who develop HUS, complicated HUS (eg, with central nervous system dysfunction) [6,76,77]. In one study, they were also associated with prolongation of gastrointestinal symptoms [6].
●Nonsteroidal anti-inflammatory drugs (including ketorolac) – These can diminish renal blood flow and, thus, can theoretically exacerbate ischemic damage to the kidney that likely underlies renal failure in HUS.
●IV ondansetron – This serotonin 5-H3 receptor antagonist can prolong the QT interval. Given that a subset of deaths during HUS are sudden, we believe that it is prudent to avoid this potentially arrhythmogenic agent [78]. Orally administered ondansetron is safer but, in our experience, rarely stops vomiting associated with STEC infections.
Other management issues
Pain control — We manage abdominal pain with boluses of isotonic fluid. This is based on our speculation that some of the pain could be related to intestinal ischemia. In particular, we avoid opioids or nonsteroidal anti-inflammatory drugs because of potential risks. (See 'Avoid potentially harmful interventions' above.)
Hypoalbuminemia — Many STEC-infected patients are hypoalbuminemic. If the patient has peripheral edema or diminishing urine output and the serum albumin concentration is <3 g/dL, we suggest judicious albumin infusions (eg, 0.5 g/kg once, and then repeated several hours later) to draw extracellular fluid into blood vessels and increase circulating blood volume. We do not administer albumin with a diuretic unless there is cardiopulmonary overload.
Anemia and thrombocytopenia — Occasionally, severe anemia from hemolysis can precede, or occur in the absence of, azotemia (ie, before criteria for HUS have been met or even in situations in which HUS does not develop). If the patient is neither tachycardic nor hypoxemic, we attempt to delay transfusion. Almost all patients with STEC-related HUS ultimately require at least one red blood cell transfusion, but delaying this intervention seems prudent because the added pigment load from continuing hemolysis might be nephrotoxic. We also monitor for hypertension during or following transfusion.
Patients infected with a high-risk STEC can also develop quite severe thrombocytopenia before, or even in the absence of, azotemia. We do not use platelet transfusions unless there is severe bleeding, which is unusual. As HUS is a thrombotic injury, clinically consequential hemorrhagic complications are rare, and platelet infusions theoretically could exacerbate the thrombotic state.
SPECIAL CIRCUMSTANCES
Exposed asymptomatic individuals — We advise individuals exposed to an outbreak source and household contacts of infected patients to monitor symptoms and seek medical care if they develop diarrhea. Otherwise, we do not recommend interventions (ie, volume expansion) in the absence of symptoms. There are no data demonstrating a benefit of antibiotics or other interventions in preventing infection or other complications in an exposed but presymptomatic individual.
Late presentation — Late presentation of STEC infection (ie, at day 9 or later, in the absence of hemolytic uremic syndrome [HUS]) is a fairly uncommon situation (algorithm 1). If patients are diagnosed with STEC in an apparent convalescent phase (ie, on or after day 10 of illness), hospital admission and intravenous fluid administration are not necessary as long as there is no evidence of HUS. We do, however, suggest that a complete blood count (CBC), blood urea nitrogen (BUN), and creatinine be checked on presentation and again the following day in such patients.
Fever or secondary infection — Many patients with STEC infection report fever prior to seeking care, but almost all are afebrile on presentation. Documentation of a fever following presentation should prompt workup for other infections; for example, we obtain blood cultures and chest radiograph.
For patients with STEC infection who have not developed HUS, the clinician should be judicious about seeking other infections and ensure that a concurrent infection truly warrants antibiotic therapy because of the association between antibiotics and HUS (see 'Antibiotics' above). It may not be necessary to treat certain coincident bacterial infections, such as otitis media, another co-detected bacterial enteric infection, or group A streptococcal pharyngitis. Although rare, STEC-infected patients can have serious simultaneous infections, such as invasive pneumococcal [50] and Clostridium septicum [79-82] infections. In these situations, antibiotics are clearly warranted.
Once a patient has developed HUS, and especially if they have developed anuric HUS, antibiotics should not be withheld if there is good justification for their use. However, we avoid ceftriaxone in such cases because HUS results in a pigment load that can lead to biliary stones [83], and ceftriaxone may exacerbate this process.
Rapidly improved symptoms — Patients whose bloody diarrhea resolves rapidly (eg, within a few hours) are not at risk of having an STEC infection; bacterial diarrhea rarely improves so rapidly. If such patients had been admitted for management, it is reasonable to discharge them after a few hours of observation.
HUS in older adults — Among older adults who develop hemolytic uremic syndrome (HUS), myocardial ischemia, acute congestive heart failure, and other cardiac complications have been reported [84]. We encourage vigilance in such patients for any signs of cardiac deterioration while hospitalized and to have a low threshold for obtaining a troponin and cardiac echocardiogram if indicated.
The approach to suspected heart failure is discussed elsewhere. (See "Heart failure: Clinical manifestations and diagnosis in adults".)
Infants — In North America, STEC infections in infants are quite rare. We therefore do not apply our fluid protocols in children younger than nine months of age prior to microbiologic confirmation of an STEC infection. There is some overlap in the clinical characteristics of STEC infections in children and intussusception (and some children can have both). (See 'Other' above.)
INFECTION CONTROL MEASURES
Inpatient precautions — Hospitalized patients with STEC infections should be placed on contact precautions to reduce person-to-person transmission [85]. (See "Infection prevention: Precautions for preventing transmission of infection", section on 'Contact precautions'.)
Public health notification — The local public health authorities should be notified of any patient diagnosed with STEC or hemolytic uremic syndrome (HUS). This is important for outbreak tracking and source identification. In the United States, the relevant authority is generally the public health department in the county in which the patient resides. The clinical microbiology laboratory or hospital infection control team can provide guidance regarding contact details.
We discourage providers from attempting to gather epidemiologic data themselves and encourage them to defer epidemiologic investigations to the public health authorities; case questionnaires used by health departments are quite detailed and are designed to gain maximal information without prompting biased responses.
Identifying sick contacts — Any acutely ill family members of patients with STEC infection should be advised to seek medical care; their clinician should be alerted to exposure to a patient with STEC infection.
Return to school or work — Local health jurisdictions have varying policies regarding exclusion from daycare and food handling jobs to prevent post-infectious transmissions of these pathogens. Physicians should defer to their policies.
A negative stool culture is often requested prior to return to daycare, school, or other sensitive settings (eg, occupation with food preparation). For children, we typically obtain a stool culture prior to discharge from the hospital, as this is more convenient than submitting an outpatient specimen.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Acute diarrhea in adults" and "Society guideline links: Acute diarrhea in children".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on “patient info” and the keyword(s) of interest.)
●Basics topic (see "Patient education: E. coli diarrhea (The Basics)")
SUMMARY AND RECOMMENDATIONS
●Terminology – Shiga toxin-producing Escherichia coli (STEC) fall into two clinically relevant categories: those that contain a gene encoding Shiga toxin 2 (with or without a gene encoding Shiga toxin 1) and those that do not (ie, their only Shiga toxin gene encodes Shiga toxin 1) (table 1). Shiga toxin 2 production is associated with serious infection characterized by bloody diarrhea and a risk for hemolytic uremic syndrome (HUS). Almost all E. coli O157:H7 contain a gene encoding Shiga toxin 2; only a subset of non-O157:H7 E. coli contain a gene encoding Shiga toxin 2. We use the term "high-risk STEC" to collectively refer to these Shiga toxin 2-producing pathogens. (See 'Terminology' above.)
●Clinical features – The course of infection with high-risk STEC follows a highly patterned sequence of events (figure 1). The first day of diarrhea is considered the first day of illness. Diarrhea is initially non-bloody, but in most patients, becomes visibly bloody one to three days later. Abdominal pain is common; fever at presentation is uncommon, although many patients report fevers early in illness (ie, before they present for care). HUS, if it occurs, is usually established between days 5 and 13 of illness. (See 'Clinical features' above.)
●Evaluation and diagnosis
•Clinical suspicion – A high-risk STEC infection should be suspected in all patients of all ages with acute bloody diarrhea and abdominal tenderness (algorithm 1). It should also be suspected in individuals with diarrhea and prior contact with a known STEC patient and in individuals with HUS, even if they do not have diarrhea. Fever does not rule out the possibility of an STEC infection. (See 'Clinical suspicion' above.)
•Microbiologic diagnosis – Rapid, accurate, and complete diagnostic microbiology is critical for detecting STEC infections. Unless the patient immediately produces a stool specimen, we suggest using rectal swab specimens to detect STEC. The diagnosis of a high-risk STEC infection is made in a patient with acute diarrhea when E. coli O157:H7 is isolated on selective culture or when Shiga toxin 2 (or a gene encoding this toxin) is detected. Although many molecular tests can distinguish between Shiga toxins 1 and 2, this detail is not necessarily reported and must be specifically requested. (See 'Microbiologic diagnosis' above.)
●Management – Our approach to managing suspected or confirmed high-risk STEC infections focuses on practices that lower the likelihood of HUS, particularly anuric HUS. We favor hospital admission for such patients to permit aggressive intravenous fluid management and daily reassessment (algorithm 1).
•Fluid management – For patients with suspected or confirmed high-risk STEC infection, we suggest early intravenous volume expansion with isotonic crystalloid with a goal of reducing the hemoglobin concentration (Grade 2C). We generally start with a bolus of at least 20 mL/kg of normal saline and administer 2 liters per m2 per day. Discontinuation of intravenous fluids depends on the stage of the illness, the platelet count, and the clinical condition based on daily monitoring. (See 'Fluid management' above.)
•Avoiding harmful interventions – We recommend not treating confirmed high-risk STEC infections with antibiotics (Grade 1B). We also do not use empiric antibiotics in suspected STEC infections, pending microbiologic testing. If STEC is identified in patients in whom antibiotic therapy was initiated empirically, we discontinue it. Antibiotics do not reduce symptoms or complications of STEC infection and have been associated with development of HUS. (See 'Antibiotics' above.)
●Infection control – Hospitalized patients with STEC infections should be placed on contact precautions to reduce person-to-person transmission. Local public health authorities should be notified. A negative stool culture is often requested prior to return to daycare, school, or other sensitive settings (eg, occupation with food preparation). (See 'Infection control measures' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Sarah Greene, MD, who contributed to a revision of this topic review.
