What it is:Acute infectious diarrhea, whether viral, bacterial, or toxin-driven, plus the special case of Clostridioides difficileinfection (CDI) that shows up after antibiotics wreck the normal gut flora.
The core problem:Whatever the bug, the thing that actually kills people is fluid and electrolyte loss, not the organism itself. Globally this is a top-5 killer of children under 5 (about 525,000 deaths a year), and in the US it disproportionately kills the elderly instead, with roughly 179 million cases, 500,000 hospitalizations, and 5,000 deaths annually.
What you do about it:Rehydrate everyone. Then decide, pathogen by pathogen, whether antibiotics help or hurt.
The single organizing question for this whole chapter is: is this a secretory (watery) process or an invasive (dysenteric) one?That split predicts the pathogen, predicts the symptoms, and predicts whether antibiotics belong in the plan at all. Get that classification right and the rest of the chapter falls into place.
Nearly every GI infection sorts into one of two clinical patterns. Learn to sort a stem into one of these columns before you even think about a drug name.
| Watery diarrhea | Dysenteric (inflammatory) diarrhea | |
|---|---|---|
| Mechanism | Enterotoxin drives active secretion; the mucosa itself is untouched | Direct bacterial invasion and damage of the gut wall |
| Stool pattern | Large volume, watery, painless | Frequent small-volume stools, often bloody, with mucus and tenesmus |
| Fever | Usually minimal or absent | Common, sometimes high with systemic toxicity |
| Classic pathogens | Norovirus, rotavirus, ETEC, V. cholerae | Campylobacter, EHEC, Salmonella, Shigella, Yersinia |
| Antibiotics indicated? | Usually not, unless severe cholera or severe/traveler's ETEC | Often yes in febrile, severe, or at-risk patients, except EHEC |
"Bloody diarrhea, so give antibiotics" is wrong for one specific bug. EHEC is the exception in the dysenteric column.Treating it with antibiotics kills the bacteria, which dumps more Shiga toxin into circulation and raises the risk of hemolytic uremic syndrome (HUS). Bloody stool plus recent undercooked beef or unpasteurized milk should make you think twice before reaching for an antibiotic, not faster.
Toxins like cholera toxin and heat-labile/heat-stable ETEC toxins hijack intestinal epithelial cells to dump chloride, sodium, and water into the lumen without ever damaging the cell itself. Because the enterocyte and its glucose-sodium cotransporter (SGLT1) are still fully intact, glucose-driven sodium (and water) absorption keeps working right alongside the toxin-driven secretion. That single fact is why oral rehydration solution works: pairing glucose with sodium in the right ratio lets the gut pull fluid back in even while the toxin is actively pushing it out.
Campylobacter, Shigella, nontyphoidal Salmonella, and EHEC all physically invade or damage the mucosa, triggering a local inflammatory response. That's why these infections produce blood, mucus, fever, and cramping instead of just volume loss, and why antibiotics have a real role here: shortening invasion shortens illness and, in shigellosis, cuts the period of fecal shedding.
Antibiotics wipe out the normal colonic flora that would otherwise keep C. difficilespores in check. Once established, toxigenic strains release toxin A and toxin B, which disrupt the colonic epithelial cytoskeleton and tight junctions, causing an inflammatory cascade that ranges from watery diarrhea to the pathognomonic pseudomembranes of pseudomembranous colitis, and in the worst cases, toxic megacolon.
Almost every therapy decision in this chapter traces back to one of two mechanisms: is fluid being secreted(fix it with ORS, an intact absorptive pathway), or is tissue being invaded(fix it, selectively, with an antibiotic)? Antimotility drugs are the wrong answer for either process when a toxin or organism needs to keep moving out of the colon rather than sit and cause more damage.
Viral gastroenteritis is the most common cause worldwide and in the US. Noroviruses alone cause more than 90% of viral gastroenteritis cases across all ages and about half of outbreaks globally.
| Virus | Peak age | Season | Duration | Notable features |
|---|---|---|---|---|
| Rotavirus | 6 mo - 2 yr | Oct - Apr | 3-7 days | Fever, vomiting, can cause secondary lactose intolerance |
| Norovirus | All ages | Winter peak | 2-3 days | Explosive, myalgias; the classic cruise-ship/outbreak bug |
| Astrovirus | <7 yr | Winter | 1-4 days | Headache, malaise alongside diarrhea |
| Enteric adenovirus | <2 yr | Year-round | 7-9 days | Respiratory symptoms overlap |
Beyond the pathogen-specific pattern, watch for the general warning signs that separate "self-limited" from "needs a workup": fever, visible blood or mucus in stool, severe abdominal pain, symptoms beyond 7 days, signs of dehydration, or diarrhea starting during or shortly after a hospital stay or antibiotic course (think CDI first in that last group).
New diarrhea in a patient who was on fluoroquinolones, clindamycin, carbapenems, or 3rd/4th-generation cephalosporinsin the last several weeks, especially if they're elderly or were recently hospitalized, is CDI until proven otherwise.
Most acute gastroenteritis is diagnosed clinically and doesn't need stool studies. Reserve stool culture, ova and parasite exam, or viral/bacterial PCR panels for diarrhea lasting more than 7 days, bloody or mucoid stool, high fever, severe dehydration, immunocompromised hosts, recent travel, or suspected outbreak. Suspected CDI is confirmed by detecting C. difficiletoxin or toxigenic organism in stool (PCR or toxin immunoassay), or by colonoscopic/histopathologic findings of pseudomembranous colitis.
Acute weight loss is the single most reliable measure of fluid deficit, but when you don't have a recent baseline weight, use the clinical exam.
| Finding | Minimal (<3% loss) | Mild-moderate (3-9% loss) | Severe (≥10% loss) |
|---|---|---|---|
| Mental status | Normal | Normal to listless | Apathetic, lethargic, comatose |
| Eyes / mouth | Normal / moist | Sunken orbits / dry | Deeply sunken / parched |
| Skin fold recoil | Normal | <2 seconds | >2 seconds |
| Pulses / heart rate | Normal | Normal to slightly ↑ | Weak/thready; may bradycardia terminally |
| Urine output | Normal to ↓ | Decreased | Minimal |
When you're not sure which bucket a patient falls into, treat for the more severe category. Undertreating dehydration is the costlier mistake.
Regardless of pathogen, fixing fluid and electrolyte losses is the first and most important intervention. Mild, self-limited cases just need oral fluids and easily digested food. Severe watery or dysenteric diarrhea needs IV rehydration, and sometimes antibiotics or antimotility agents on top.
ORS works because it pairs glucose with sodium, exploiting the SGLT1 cotransporter that keeps functioning even during active toxin-driven secretion (see Pathophysiology). Give it in small, frequent volumes, about 5 mL every 2-3 minutes by spoon or oral syringe, since large volumes at once can worsen vomiting.
| Solution | Na (mEq/L) | K (mEq/L) | Carbohydrate (mmol/L) | Notes |
|---|---|---|---|---|
| WHO/UNICEF (2002) | 75 | 20 | 75 | Reference standard |
| Pedialyte | 45 | 20 | 140 | Common OTC option |
| Rehydralyte | 75 | 20 | 140 | Higher-sodium rehydration product |
| Avoid these for actual rehydration | ||||
| Apple juice / sports drinks / chicken broth | Wrong Na:carbohydrate ratio | Too much sugar, too little sodium; can worsen osmotic diarrhea | ||
Patients (and well-meaning parents) reach for sports drinks, apple juice, or broththinking any fluid will do. All three have the wrong sodium-to-carbohydrate ratio for true rehydration and can make osmotic diarrhea worse. ORS is not interchangeable with "drink more fluids."
Most acute gastroenteritis, especially viral, resolves without antibiotics. Reserve antimicrobials for: severe diarrhea, moderate-to-severe traveler's diarrhea, most febrile dysenteric diarrhea, and culture-confirmed bacterial diarrhea. In dysenteric disease specifically, antibiotics are typically reserved for at-risk groups: the elderly, immunocompromised patients, daycare-attending children, malnourished children, and healthcare workers, since treating everyone doesn't change the overall course much and drives resistance.
Never treat suspected or confirmed EHEC with antibiotics.Bacterial killing releases more Shiga toxin and increases the risk of hemolytic uremic syndrome. Antibiotics are, however, appropriate and beneficial in severe cholera and severe ETEC diarrhea, the two watery-diarrhea exceptions to "watery usually doesn't need antibiotics."
| Pathogen | Pediatric | Adult |
|---|---|---|
| Watery diarrhea | ||
| ETEC | Azithromycin 10 mg/kg/day IV/PO once daily × 3 days; or ceftriaxone 50 mg/kg/day IV once daily × 3 days | Ciprofloxacin 750 mg PO once daily × 1-3 days; alt: rifaximin 200 mg PO TID × 3 days, or azithromycin 1000 mg × 1 or 500 mg daily × 3 days |
| Vibrio choleraeO1 | Erythromycin 30 mg/kg/day divided q8h × 3 days; or azithromycin 10 mg/kg/day once daily × 3 days | Doxycycline 300 mg PO × 1 dose; alt: azithromycin 500 mg daily × 3 days, ciprofloxacin 750 mg daily × 3 days, or IV ceftriaxone |
| Dysenteric diarrhea | ||
| Campylobacter | Azithromycin 10 mg/kg/day × 3-5 days; or erythromycin 30 mg/kg/day divided BID-QID × 3-5 days | Azithromycin 500 mg daily × 3 days; alt: ciprofloxacin 750 mg daily × 7 days |
| Salmonella(nontyphoidal, high-risk only) | Ceftriaxone 100 mg/kg/day divided q12h × 7-10 days; or azithromycin 20 mg/kg/day × 7 days | Ceftriaxone 2 g IV/IM × 1; or ciprofloxacin 750 mg daily × 7-10 days (14 days if immunocompromised) |
| Shigella(high-risk only) | Azithromycin 10 mg/kg/day × 3 days; or ceftriaxone 50 mg/kg/day × 3 days | Azithromycin 500 mg daily × 3 days; ceftriaxone 2 g IV/IM × 1; or ciprofloxacin 750 mg daily × 3 days |
| Yersinia(high-risk only) | Treat as pediatric shigellosis | Trimethoprim-sulfamethoxazole 160/800 mg BID × 7 days; alt: cefotaxime IV or ciprofloxacin 750 mg daily × 7 days |
Azithromycin and ceftriaxone show up as first-line or alternative options across almost every pathogen in this table. If you forget a specific regimen on an exam, those two are usually a defensible guess for empiric dysenteric bacterial diarrhea (except EHEC, where the answer is still no antibiotic at all).
The syndrome travelers know well: malaise, anorexia, and cramping followed by sudden-onset diarrhea. Caused by contaminated food or water, most commonly ETEC, plus Campylobacter, Shigella, and Salmonella.
C. difficileis an anaerobic, spore-forming, gram-positive rod, and the most commonly recognized cause of healthcare-associated infectious diarrhea. Spores survive on surfaces and hands, which is why alcohol-based hand sanitizer alone doesn't reliably clear them (soap and water does the mechanical job better in an active outbreak).
Clindamycin carries a black box warning for CDIand roughly a 20-fold increased odds of CDI compared to no antibiotic exposure, well above the roughly 5-fold odds ratio seen with most other antibiotic classes. Fluoroquinolones, carbapenems, and 3rd/4th-generation cephalosporins are the other major offenders. This is the single most testable CDI fact from lecture.
| Severity | Markers | Treatment |
|---|---|---|
| Nonsevere | WBC ≤15,000 cells/mm³, SCr <1.5 mg/dL | Vancomycin 125 mg PO QID × 10 days, OR fidaxomicin 200 mg PO BID × 10 days (metronidazole 500 mg PO q8h × 10 days only if the others are unavailable/unaffordable) |
| Severe | WBC >15,000 cells/mm³, SCr >1.5 mg/dL | Vancomycin 125 mg PO QID × 10 days, OR fidaxomicin 200 mg PO BID × 10 days |
| Fulminant | Hypotension/shock, ileus, or megacolon | Metronidazole 500 mg IV q8h PLUSvancomycin 500 mg q6h via NG or PO (rectally if ileus present) |
Older teaching made metronidazole first-line for mild CDI. Current SHEA/IDSA guidance makes oral vancomycin or fidaxomicin first-line for both nonsevere and severe disease, with fidaxomicin preferredoverall. Metronidazole is now a fallback, used PO/IV in fulminant disease specifically as an add-on to high-dose vancomycin, not as monotherapy for routine cases. Treatment courses run 10 days, and repeat stool testing is not used as a test of cure since patients can remain colonized without active disease.
Recurrence happens in 25-35% of patients. Management depends on what was used initially and how many recurrences have occurred:
Supportive carealways includes fluid/electrolyte replacement and stopping the offending antibiotic when possible. Antimotility agents and bile-acid-binding resins (cholestyramine, colestipol) have been tried in CDI but are discouraged, since slowing transit and binding the antibiotic itself can worsen outcomes.
Worth repeating because it's the highest-yield "don't" in this chapter: never give antibiotics for suspected EHEC.The clinical clue is bloody diarrhea with a history of undercooked ground beef, unpasteurized dairy/juice, or contaminated produce, especially in a child.
Avoid loperamide and diphenoxylate/atropine in EHEC, shigellosis, pseudomembranous colitis (CDI), and febrile dysenteric diarrhea generally.Trapping toxin and organisms in the colon can precipitate toxic megacolon.
| Parameter | When | Watching for |
|---|---|---|
| Weight | Baseline and through rehydration | Most reliable marker of fluid deficit and of resolution |
| Mental status, skin turgor, urine output | Every reassessment during rehydration | Ongoing dehydration despite therapy |
| Stool frequency/consistency | Daily | Response to therapy; new blood/mucus signals a change in diagnosis |
| WBC, SCr | At CDI diagnosis and through treatment | Severity classification (nonsevere vs severe vs fulminant) and treatment response |
| Electrolytes (Na, K) | Through IV rehydration, more often if severe | Correction of losses without overcorrection |
| Symptom duration | Ongoing | >48 hours on loperamide, or >7 days overall, warrants re-evaluation |
| Recurrence symptoms post-CDI treatment | Weeks after completing therapy | New diarrhea suggesting relapse (no test-of-cure stool testing needed) |