Peritonitis in ICU: spontaneous bacterial, secondary, and tertiary

Cinematic clinical photograph of cloudy peritoneal dialysis effluent in a sample pot beside a septic patient monitor, ICU setting, clinical-blue lighting, no text, no people — FigureSecondary and tertiary peritonitis demand surgical source control — antibiotics alone fail.

Peritonitis classification: spontaneous bacterial peritonitis, secondary peritonitis from perforation or abscess, tertiary peritonitis with resistant organisms — educational clinical infographic — FigureThree peritonitis types — SBP (ascitic PMN 250 or more without surgical source), secondary (source control mandatory), tertiary (persistent, resistant flora).

Pathophysiology of peritonitis: bacterial translocation in cirrhosis, hollow viscus perforation, cytokine cascade and septic shock — educational diagram — FigureSBP follows bacterial translocation into low-opsonin ascites; secondary peritonitis follows hollow-viscus perforation or anastomotic leak.

Peritonitis management: SBP antibiotics plus albumin day 1 and 3, secondary peritonitis source control within 6 hours, broad empiric cover for tertiary disease — clinical-blue pathway — FigureSBP: third-generation cephalosporin plus albumin. Secondary: source control plus broad antibiotics. Tertiary: resistant-organism cover and re-look decisions.

Types of peritonitis

Feature	SBP	Secondary	Tertiary	CAPD
Source	None (spontaneous)	Perforation, leak, abscess	Persistent/recurrent after secondary	Peritoneal dialysis catheter
Population	Cirrhosis with ascites	Anyone with abdominal pathology	Critically ill ICU patients	PD patients
Organisms	Monomicrobial (E. coli, Klebsiella)	Polymicrobial (gram-negative, anaerobes)	Resistant (Enterococcus, Candida, Pseudomonas, MRSA)	Staphylococcus (epidermidis, aureus)
Diagnosis	Ascitic PMN >250/mm³	CT, laparotomy	Persistent sepsis, cultures positive	Cloudy dialysate, PMN >100/mm³
Treatment	Cefotaxime + ALBUMIN	SURGERY (source control) + broad antibiotics	Anti-fungal, anti-Enterococcus, anti-Pseudomonas	Intraperitoneal antibiotics ± catheter removal
Mortality	10-30%	5-20%	30-50%	<5% (per episode)

[1]

Management of suspected secondary peritonitis

Resuscitate — ABC. IV fluids (septic patients often hypovolaemic). Vasopressors (noradrenaline) for shock. Broad-spectrum antibiotics within 1h
CT abdomen with contrast — identify source (perforation, abscess, leak, ischaemia)
Source control — SURGERY is MANDATORY: (a) Perforation — repair/resect. (b) Abscess — drain (percutaneous or surgical). (c) Leak — repair/re-anastomose/stoma. (d) Ischaemia — resect necrotic bowel. (e) Lavage (wash out peritoneum)
Antibiotics — broad-spectrum, polymicrobial cover: piperacillin/tazobactam OR meropenem + metronidazole. Cover gram-negative, anaerobes, Enterococcus. Consider antifungal (candida) if prolonged, immunosuppressed
ICU support — ventilation, renal support, vasopressors
Duration: 4 days after source control (STOP-IT trial showed no benefit of longer)
Monitor for tertiary peritonitis — persistent sepsis despite source control → resistant organisms → broaden antibiotics

[1] [1]

SAQ — Spontaneous bacterial peritonitis in a cirrhotic with decompensation

10 minutes · 10 marks

A 54-year-old man with known alcohol-related cirrhosis (Child-Pugh C, MELD-Na 22) presents with 36 hours of abdominal pain, fever 38.6 degrees C, and increasing confusion. He is on no SBP prophylaxis. On arrival he is drowsy (GCS 13), BP 88/54 (MAP 65), HR 112, RR 22, SpO2 96 per cent on room air, T 38.4 degrees C. The abdomen is distended with a positive fluid thrill and generalised mild tenderness without guarding. Blood tests: Hb 96, WCC 14.2, neutrophils 11.8, platelets 78, creatinine 138 micromol/L (baseline 92), bilirubin 92 micromol/L, INR 1.9, albumin 22 g/L, lactate 1.6. Diagnostic paracentesis returns turbid fluid; the bedside dipstick is leukocyte esterase 3+.

[1]

SAQ — Secondary peritonitis from a perforated viscus with septic shock

10 minutes · 10 marks

A 67-year-old man presents with 14 hours of severe generalised abdominal pain and vomiting. He is on apixaban for non-valvular atrial fibrillation. On arrival he is clammy and drowsy (GCS 14), temperature 38.9 degrees C, HR 132 in AF, BP 78/42 (MAP 54) on 5 L/min nasal spec, RR 28, SpO2 94 per cent. The abdomen is rigid and distended with generalised peritonism and absent bowel sounds. Bloods: Hb 121, WCC 22.4, lactate 5.8, creatinine 188 micromol/L, venous pH 7.21, base excess -10, INR 1.4. An erect chest x-ray shows free gas under the diaphragm. He has had no recent surgery; he takes apixaban 5 mg BD with last dose 8 hours ago.

[1]

Clinical pearls

High-yield peritonitis points for CICM/FFICM exam

SBP = ascitic PMN >250 cells/mm³. Diagnostic paracentesis is MANDATORY in any cirrhotic with ascites and new fever/abdominal pain/altered mental state/renal failure. DO NOT wait for culture (50% culture-negative) — treat based on PMN count.[1] }
SBP organisms are usually MONOMICROBIAL (unlike secondary peritonitis which is polymicrobial). Gram-negative rods (E. coli 70%, Klebsiella) most common. Gram-positive (Streptococcus pneumoniae, Enterococcus) in 25%. Anaerobes RARE (protected by low ascitic oxygen).[1] }
Albumin in SBP — reduces hepatorenal syndrome and mortality. Landmark trial (Sort 1999, NEJM): SBP patients given cefotaxime + albumin vs cefotaxime alone. Albumin REDUCED HRS (10% vs 33%) and mortality (10% vs 29%). Dose: 1.5 g/kg day 1, 1 g/kg day 3. Give to ALL SBP patients (especially with risk factors: bilirubin >68, creatinine >88).[2] }
SBP treatment: cefotaxime 2g IV BD-TDS for 5-7 days. Alternative: ceftriaxone 2g IV OD. Oral ofloxacin (400mg BD) for uncomplicated SBP if not septic and no prior antibiotic prophylaxis. 5 days is as effective as 10 days.[1] }
SBP prophylaxis: norfloxacin 400mg OD or ciprofloxacin 500mg OD. Indications: (1) Previous SBP (lifelong prophylaxis). (2) Ascitic protein <15 g/L (especially if also renal/liver dysfunction). (3) GI bleed in cirrhotic (prevents SBP and bacterial translocation). Primary prophylaxis controversial.[1] }
Secondary peritonitis: source control is MANDATORY. Antibiotics alone CANNOT cure secondary peritonitis — the intra-abdominal source (perforation, leak, abscess) must be addressed surgically. Delay in source control >24h increases mortality significantly.[3] }
Secondary peritonitis is POLYMICROBIAL. Gram-negative (E. coli, Klebsiella), anaerobes (Bacteroides, Clostridium), Enterococcus. Antibiotics: piperacillin/tazobactam OR meropenem + metronidazole. Cover Enterococcus (especially post-operative, healthcare-associated).[3] }
STOP-IT trial (2015): 4 days antibiotics after source control. RCT: complicated intra-abdominal infection. 4 days vs ~8 days after source control. Result: NO difference in outcomes (infection-related adverse events, mortality). SHORT course (4 days after adequate source control) is sufficient.[4] }
Tertiary peritonitis: persistent/recurrent after treatment of secondary. Occurs 30-50% of patients after secondary peritonitis. Organisms: RESISTANT — Enterococcus (VRE), Candida, Pseudomonas, MRSA, Stenotrophomonas. Reflects immune dysfunction of critically ill. Mortality 30-50%. Treatment: broad-spectrum (vancomycin, meropenem, echinocandin/fluconazole), anti-fungal essential.[5] }
Candida peritonitis. Common in: post-operative, prolonged antibiotics, immunosuppressed, multiple abdominal surgeries, severe pancreatitis. Treatment: fluconazole (if susceptible) or echinocandin (caspofungin, micafungin) if critically ill/unstable. Remove infected foreign material. Source control essential.[3] }
CAPD peritonitis: cloudy dialysate + abdominal pain. Diagnosis: dialysate WBC >100/mm³ (after dwell), PMN >50%. Organisms: coagulase-negative Staph (40% — from touch contamination), Staph aureus (15% — exit-site infection), gram-negative (15%), fungal (2% — needs catheter removal). Treatment: INTRAPERITONEAL antibiotics (vancomycin/cefazolin + ceftazidime/gentamicin). Catheter removal if: fungal, refractory, recurrent, Pseudomonas.[6] }
Post-operative peritonitis (anastomotic leak). Presents day 3-7 post-surgery with fever, ileus, abdominal pain, sepsis, drain output (faeculent/bilious). CT with water-soluble contrast (water-soluble — safer than barium if leak). Management: re-operation (repair/defunction), antibiotics, ICU.[3] }
Abdominal sepsis is a common cause of septic shock in ICU. Always consider intra-abdominal source in septic patient without obvious cause. CT abdomen is essential. Source control (surgery/drainage) is as important as antibiotics. Procalcitonin may help (elevated in bacterial infection).[3] }
Diagnostic peritoneal lavage (DPL) — largely replaced by FAST ultrasound and CT. Still used occasionally in trauma (if FAST negative but suspicion high). Positive: >100,000 RBC/mL (indicating hollow viscus injury requiring laparotomy) or bile/food/faeces (bowel injury).[3] }

Red flags

Prognosis

Albumin in SBP (Sort 1999, NEJM)

RCT: 126 cirrhotic patients with SBP. Cefotaxime + albumin vs cefotaxime alone.

Hepatorenal syndrome: 10% (albumin) vs 33% (control) — p=0.002, 67% relative reduction
In-hospital mortality: 10% (albumin) vs 29% (control) — p=0.01, 65% relative reduction
3-month mortality: 22% vs 41% — p=0.03 [1]

Albumin dose: 1.5 g/kg day 1, 1 g/kg day 3. CONCLUSION: Albumin is STANDARD OF CARE for SBP (especially with bilirubin >68 μmol/L or creatinine >88 μmol/L).

SBP — spontaneous bacterial peritonitis: deep dive [1]

Pathophysiology and portal of entry — why the four peritonitides differ

Type	Source / portal of entry	Microbiology	Key host factor
SBP	Haematogenous / lymphatic seeding of ascites; intestinal bacterial translocation across oedematous, more-permeable cirrhotic gut wall. NO visible intra-abdominal source on imaging.	Monomicrobial — single organism in 90%+ of culture-positive cases	Cirrhosis with ascites: low ascitic opsonic/complement activity (protein <15 g/L, C3 low) impairs phagocytosis. Diseased Kupffer cells fail to clear portal bacteraemia.
Secondary	Spillage from a breached viscus — perforation, ischaemic necrosis, anastomotic leak, abscess, trauma. Source is anatomical and IMAGING-DETECTABLE.	Polymicrobial — gut flora: gram-negative bacilli, anaerobes (Bacteroides, Clostridium), Enterococcus, streptococci; ± Candida	Any patient with abdominal pathology; severity depends on load, virulence, host, and source-control TIMING.
Tertiary	Persistent or recurrent infection after apparent control of secondary peritonitis — failure of source control or host immunity. Often NO clear new source.	MDR/fermenters/fungi — Enterococcus (incl. VRE), Candida, Pseudomonas, Stenotrophomonas, MRSA, Acinetobacter	Critically ill ICU patient: immune paralysis, prolonged antibiotics, foreign material (mesh, drains), open abdomen, TPN. Reflects immunoparesis more than new contamination.
CAPD / catheter-related	Inoculation along the Tenckhoff catheter (touch contamination, exit-site/tunnel infection) or trans-visceral migration. Peritoneum is the dialyser.	Staphylococci (CoNS 30-45%, S. aureus 10-20%), gram-negatives (15-25%), Streptococcus (10%), fungal (2-5%)	Peritoneal dialysis patients; defence diluted by ~2 L dialysate, low opsonin, catheter as foreign body.

[1]

Ascitic fluid analysis — interpreting the diagnostic paracentesis

Result	Definition	Significance	Action
SBP	Ascitic PMN ≥ 250 cells/mm³ (with or without positive culture)	Diagnostic of SBP — treat empirically NOW. Culture positive in only ~50%.	Cefotaxime 2 g IV BD-TDS × 5-7 days + albumin 1.5 g/kg day 1, 1 g/kg day 3.[1]
Culture-negative neutrocytic ascites (CNNA)	PMN ≥ 250/mm³ AND negative culture (and no antibiotics in prior 7 days)	A variant of SBP — identical prognosis and management. The 50% culture-negative group.	Treat as SBP. Not a false positive — usually due to culture technique; bedside inoculation into blood-culture bottles ↑ yield to 70-80%.[1]
Monomicrobial non-neutrocytic bacterascites (MNB)	Single organism on culture AND PMN < 250/mm³	Often transient/colonisation that self-clears; BUT treat if symptomatic (fever, pain) or cirrhotic — risk of progression to SBP.	Treat if symptomatic; repeat tap in 24-48 h if asymptomatic.
Polymicrobial bacterascites	Multiple organisms AND PMN < 250/mm³	Suggests PERFORATION — bowel leak inoculating ascites. Diagnostic clue that this is NOT SBP.	Image (CT) for perforation → surgical source control.
Secondary peritonitis (in cirrhotic with ascites)	PMN ≥ 250 AND polymicrobial culture (multiple organisms, including anaerobes/yeast) AND ascitic: total protein >10 g/L, glucose <2.8 mmol/L, LDH >225 U/L	Differentiate from SBP: SBP is monomicrobial, no anaerobes, normal glucose/LDH. Secondary needs surgery.	CT + surgery; broaden to cover anaerobes + Enterococcus.

Diagnostic workup of a cirrhotic patient with ascites and fever/altered mental state

Diagnostic paracentesis is MANDATORY — every cirrhotic with ascites + any of: fever/abdominal pain/altered mental state/renal impairment/GI bleed/leucocytosis. Also tap on admission and at first sign of decompensation. Up to 30% of asymptomatic cirrhotic inpatients have SBP on admission.[1]
Bedside technique — supine, LLQ or RLQ (midline 2 cm below umbilicus in obese). Clean skin. 20-30 mL aspirated. Inoculate blood-culture bottles directly at the bedside (aerobic + anaerobic) — doubles culture yield vs delayed lab inoculation. Send: cell count + differential, protein, albumin, glucose, LDH, Gram stain, culture.
Reagent strip (dipstick) PMN — a leukocyte esterase dipstick (≥2+) gives a rapid bedside PMN surrogate when the lab count is delayed. Sensitivity for SBP ~90%.
Diagnosis = PMN ≥ 250 cells/mm³ — start antibiotics immediately; do NOT wait for culture. If PMN borderline (e.g., 200-250), repeat in 24-48 h.
Coagulopathy is NOT a contraindication — cirrhotic coagulopathy/ascites do not require fresh frozen plasma/platelets before a therapeutic or diagnostic tap (mild bleeding risk <1%, life-threatening <0.1%).[1]
If PMN ≥ 250 → CT abdomen/pelvis if any feature of secondary peritonitis (polymicrobial Gram stain, anaerobes, ascitic glucose <2.8 or LDH >225, free gas on CXR/CT) to exclude a perforated viscus.
Reassess at 48 h — repeat tap. PMN should fall by ≥ 25% at 48 h with appropriate antibiotic. If not, suspect resistant organism or secondary peritonitis → broaden + image.

Organisms in SBP — predicted microbiology

Group	Frequency	Organisms	Antibiotic implication
Gram-negative bacilli	70-75%	Escherichia coli (most common, ~50%), Klebsiella pneumoniae, other Enterobacterales	Third-generation cephalosporin (cefotaxime/ceftriaxone) — covers ~95% of community-acquired SBP.
Gram-positive cocci	20-25%	Streptococcus pneumoniae (pneumococcus), other streptococci, Enterococcus (more common nosocomial)	Cefotaxime covers streptococci/pneumococcus; Enterococcus needs ampicillin/vancomycin (cover if nosocomial).
Anaerobes	<5% (RARE)	Bacteroides, Clostridium	The low ascitic oxygen tension + presence of complement selects against anaerobes — do NOT add metronidazole routinely for classic SBP.
Nosocomial / healthcare-associated SBP	rising	MDR gram-negatives (ESBL, KPC), Enterococcus (incl. VRE), Pseudomonas	Switch to piperacillin/tazobactam, meropenem, or ceftolozane/tazobactam based on local antibiogram. Cefotaxime fails in ~20% of nosocomial SBP.[3]
Fungal	<2%	Candida	Only in heavily antibiotic-exposed/immunosuppressed — add echinocandin.

SBP antibiotic therapy — empiric regimens by clinical context

Clinical context	First-line	Alternative / allergy	Duration
Community-acquired SBP, uncomplicated	Cefotaxime 2 g IV BD-TDS OR ceftriaxone 2 g IV OD	Levofloxacin 500 mg IV/PO OD or moxifloxacin (if no prior quinolone prophylaxis, no vomiting/shock)	5 days (as effective as 10 days)
SBP with sepsis/septic shock	Cefotaxime/ceftriaxone ± piperacillin/tazobactam (broaden for healthcare acquisition)	Meropenem 1 g IV TDS	7 days; reassess at 48 h by PMN fall
Healthcare-associated / nosocomial SBP	Piperacillin/tazobactam 4.5 g IV TDS-QDS OR meropenem	Add vancomycin if VRE/Enterococcus risk; consider ceftolozane/tazobactam for MDR Pseudomonas	7 days, de-escalate on culture
Quinolone prophylaxis failure	NEVER use ofloxacin/ciprofloxacin empirically — high quinolone resistance	Cefotaxime/ceftriaxone (1st)	5-7 days
Add ALBUMIN to ALL SBP	1.5 g/kg IV day 1, 1 g/kg day 3	(Saline is NOT a substitute — must be albumin)	Days 1 and 3 only

[1]

SBP management algorithm — bedside sequence

Confirm diagnosis — ascitic PMN ≥ 250 cells/mm³.
Cefotaxime 2 g IV BD-TDS (or ceftriaxone 2 g IV OD) — start within 1 h of tap. Send blood cultures (bacteraemia in 30-50%).
Albumin 1.5 g/kg IV on day 1, 1 g/kg IV on day 3 — give to ALL SBP patients; MANDATORY if any of: creatinine >88 μmol/L, BUN >30 mg/dL, bilirubin >68 μmol/L.[2]
Re-tap at 48 h — PMN must fall ≥ 25%. If not: suspect resistant organism or secondary peritonitis → broaden (piperacillin/tazobactam or meropenem), CT to exclude perforation.
Search for and treat precipitant — GI bleed (give SBP prophylaxis), urinary/biliary infection, prior paracentesis without albumin, dehydration, large-volume paracentesis without albumin replacement.
Continue therapy 5 days total for uncomplicated SBP; 7 days for sepsis/bacteraemia.
Start SECONDARY PROPHYLAXIS at discharge — norfloxacin 400 mg OD or ciprofloxacin 500 mg OD (lifelong, until transplant). Recurrence 70% at 1 year without prophylaxis vs 20% with.[1]
Refer for liver transplant evaluation — SBP is a decompensating event that doubles 1-year mortality (30-50%). MELD-Na drives listing priority.

Albumin in cirrhosis — three landmark trials, three different conclusions

Trial	Population	Intervention	Key result	Interpretation
Sort 1999 (NEJM)	126 SBP patients	Cefotaxime + albumin (1.5 g/kg D1, 1 g/kg D3) vs cefotaxime alone	HRS 10% vs 33%; hospital mortality 10% vs 29%	Albumin is STANDARD OF CARE for SBP.[2]
ANSWER 2018 (Lancet)	440 decompensated cirrhosis (no SBP)	Long-term weekly albumin × 18 months vs standard care	50% reduction in SBP, fewer decompensation events, 18-month survival improved (74% vs 62%)	Long-term albumin reduces SBP incidence and improves survival in selected outpatients.[8]
ATTIRE 2021 (NEJM)	777 hospitalised cirrhosis patients	Targeted 20% albumin infusions to keep serum albumin >30 g/L vs standard care	No reduction in new infection, renal dysfunction, or mortality	Albumin infusions to a target level in unselected inpatients do NOT prevent infection. Be selective — benefit proven in SBP and ANSWER-selected outpatients only.[9]

ANSWER trial — Caraceni 2018, Lancet (long-term albumin in decompensated cirrhosis)

Open-label RCT, Italy. 440 patients with decompensated cirrhosis, no SBP at entry.

Intervention: human albumin 1.5 g/kg/week for 2 weeks, then 40 g/week for 18 months, vs standard care.
Primary endpoint — 18-month mortality: 77/220 (albumin) vs 99/220 (standard) — HR 0.66, borderline.
Incident SBP: 10% vs 24% — hazard ratio 0.40 (60% reduction).
Grade 3-4 ascites, renal dysfunction, bacterial infection: all significantly reduced.
CONCLUSION: In selected outpatients with decompensated cirrhosis, long-term albumin reduces SBP incidence and may improve survival. Selective use, not blanket ATTIRE-style infusions.[8]

ATTIRE — China 2021, NEJM (albumin in hospitalised cirrhosis)

Multicentre RCT, UK. 777 hospitalised patients with cirrhosis and serum albumin <30 g/L.

Intervention: 20% human albumin to maintain serum albumin ≥30 g/L (target-led) up to 14 days vs standard care.
Primary composite (new infection, renal dysfunction, death) at 15 days: 29.7% vs 30.2% — no difference.
Adverse events: pulmonary oedema more common with albumin.
CONCLUSION: Routine albumin infusions in unselected hospitalised cirrhosis do NOT improve outcomes. Albumin should be reserved for SBP, large-volume paracentesis (≥5 L), HRS, and selected ANSWER-style outpatients.[9]

Secondary peritonitis: deep dive — sources, organisms, source control

Sources of secondary peritonitis — predicted by anatomical site of perforation

Source	Classic organism profile	Clinical clues	Imaging
Perforated appendicitis	E. coli, Bacteroides fragilis, Streptococcus anginosus group	Migration of periumbilical→RLQ pain; McBurney's point tenderness; young male; perforation → diffuse peritonitis/sepsis	CT: peri-appendiceal fat stranding, free fluid ± free gas, appendicolith
Perforated diverticulitis (Hinchey III-IV)	E. coli, Klebsiella, Bacteroides, Clostridium; ± Enterococcus	LLQ pain in older patient; acute severe abdominal pain + rigid abdomen in Hinchey IV; faeculent peritonitis	CT: sigmoid diverticula, free gas ± faeces; Hinchey stage
Perforated peptic ulcer (PUD)	Streptococcus, E. coli, anaerobes (after several hours)	Sudden "tearing" epigastric pain → rigid abdomen; "Valentino's sign" (RIF pain from gastric fluid tracking down right paracolic gutter); standing CXR: free subdiaphragmatic gas	Erect CXR; CT abdomen; "Valsalva/computed tomography" shows anterior free gas
Ischaemic bowel / mesenteric infarction	E. coli, Klebsiella, Clostridium perfringens, Enterococcus; high risk of bacteraemia	AF, heart failure, vasopressors; "pain out of proportion"; metabolic acidosis, rising lactate; bloody diarrhoea late	CT: pneumatosis intestinalis, portal venous gas, mesenteric vessel occlusion, thickened oedematous loops
Anastomotic leak (postoperative, day 3-7)	Enterococcus (ampicillin-resistant), Candida, Pseudomonas, MDR GNR — reflects hospital flora	Fever, ileus, drain output (faeculent/bilious/purulent), sepsis after GI surgery	CT with water-soluble oral/rectal contrast (NOT barium); contrast extravasation
Biliary source (cholecystitis, cholangitis, bile leak)	E. coli, Klebsiella, Enterococcus, Enterobacter; less often anaerobes (Clostridium perfringens — emphysematous)	RUQ pain, jaundice, Charcot's/Reynold's pentad; post-cholecystectomy/ERCP; bile-stained peritoneal fluid	MRCP/ERCP; CT: wall thickening, pericholecystic fluid, ductal dilatation
Post-traumatic (blunt or penetrating bowel injury)	E. coli, Bacteroides, Streptococcus	Mechanism (seatbelt sign, handlebar, GSW); delayed presentation; persistent tachycardia/lactataemia after trauma	CT: free gas, free fluid, bowel wall thickening, mesenteric stranding

[1]

Predicted organisms and antibiotic cover by source (community vs healthcare-associated)

Severity / context	Predicted cover required	Empiric regimen
Community-acquired, mild-moderate	Gram-negative rods + anaerobes	Ceftriaxone 2 g IV OD + metronidazole 500 mg IV TDS, OR co-amoxiclav 1.2 g IV TDS + metronidazole, OR moxifloxacin 400 mg IV OD (monotherapy)
Community-acquired, high-severity / sepsis	Gram-negative rods + anaerobes + Enterococcus	Piperacillin/tazobactam 4.5 g IV TDS-QDS, OR meropenem 1 g IV TDS + metronidazole (if ESBL risk)
Healthcare-associated / postoperative	Gram-negatives (incl. ESBL/Pseudomonas) + Enterococcus + Candida	Piperacillin/tazobactam OR meropenem + vancomycin (Enterococcus/MRSA) + echinocandin (caspofungin 70 mg then 50 mg OD) if recent fluoroquinolone/cephalosporin exposure, TPN, prolonged ICU stay
Biliary source	Gram-neg + Enterococcus (anaerobes RARE in pure biliary disease unless emphysematous)	Ceftriaxone + metronidazole; piperacillin/tazobactam monotherapy if severe
Candida peritonitis (post-op, multiple abdominal re-explorations, necrotising pancreatitis, immunosuppressed)	C. albicans, C. glabrata, C. krusei	Echinocandin (caspofungin/micafungin/anidulafungin) if critically ill; step down to fluconazole if stable and C. albicans susceptible. Remove infected foreign material; source control.[10]

Principles of SOURCE CONTROL — the four D's

DRAINAGE — percutaneous (image-guided) or surgical drainage of collections/abscess. Aim: evacuating pus and contaminated fluid.
DEBRIDEMENT — excision of necrotic, ischaemic or grossly contaminated tissue (non-viable bowel, omental slough). Necrotic tissue is a culture medium that antibiotics cannot penetrate.
DECOMPRESSION / DIVERSION — relieve obstruction (NGT, rectal tube, defunctioning stoma); prevent ongoing contamination.
DEFINITIVE REPAIR — close or resect the breach: primary repair, resection + anastomosis, resection + stoma (Hartmann's), or controlled fistula formation (open abdomen with VAC, "planned re-laparotomy"). [1]

Timing: Source control within 6 h of septic shock onset is associated with reduced mortality (Surviving Sepsis). Each additional hour of delay in the hypotensive patient increases mortality ~7-8%.[11]

Management of suspected perforated viscus with septic shock

ABCDE resuscitation — oxygen, two large-bore IVs. Crystalloid 30 mL/kg bolus for sepsis-induced hypoperfusion. Noradrenaline first-line vasopressor if MAP <65 mmHg after fluids. Add vasopressin / adrenaline if refractory.
Broad-spectrum antibiotics within 1 h (sepsis hour-1 bundle) — piperacillin/tazobactam 4.5 g IV ± vancomycin (Enterococcus/MRSA) ± echinocandin if Candida risk. Add metronidazole if using ceftriaxone/cefepime alone.
Stat surgical review + anaesthetics — declare a surgical emergency; consent for laparotomy.
Cross-match 4-6 units, group-and-save, check coagulation, platelets, lactate, venous gas, lactate trend.
CT abdomen/pelvis with IV + water-soluble oral/rectal contrast — only if the patient can be stabilised (CT is inadvisable in extremis — go straight to theatre). Confirm free gas, perforation site, abscess.
Urinary catheter + arterial line + CVC — guide fluid responsiveness and vasopressors; target MAP ≥65, lactate clearance ≥10%/h, ScvO₂ ≥70%.
Theatre — emergency laparotomy: peritoneal lavage, identify and repair/resect source, drain. Damage-control surgery in unstable patients (leave abdomen open, VAC, planned second-look).
Post-op ICU — vasopressor wean, lung-protective ventilation, renal support, glycaemic control 6-10 mmol/L, deep vein thrombosis prophylaxis once bleeding controlled, stress ulcer prophylaxis, enteral nutrition within 24-48 h.
Antibiotic duration: 4 days after adequate source control (STOP-IT).[4] Prolong if persistent sepsis, bacteraemia, fungaemia, source control incomplete.

STOP-IT trial — Sawyer 2015, NEJM (duration of antibiotics after source control)

Multicentre RCT, USA. 518 patients with complicated intra-abdominal infection and adequate source control.

Intervention: antibiotics stopped after a fixed ~4-day course (approx. 96 h) vs continued until physician-satisfied (~8 days).
Primary composite (surgical-site infection, recurrent IAI, death) at 30 days: 21.8% (short) vs 22.3% (long) — non-inferior.
Median duration: 4 days (short) vs 8 days (long).
CONCLUSION: Once adequate source control is achieved, ~4 days of antibiotics is sufficient. Prolonged courses do not improve outcomes and select for resistance/C. difficile.[4]

Tertiary peritonitis: deep dive — the failing host

Secondary vs tertiary peritonitis — distinguishing the ICU phenotype

Feature	Secondary peritonitis	Tertiary peritonitis
Timing	Acute — first presentation of perforation/leak	Persistent or recurrent after ≥48 h of treatment of secondary, OR no source identifiable
Host	Often previously well; new intra-abdominal catastrophe	Critically ill; prolonged ICU stay, multi-organ failure, immunoparesis
Source	Identifiable perforation/leak/abscess	Often NO source found at re-operation; widespread serosal inflammation, "sticky" fibrinous peritoneum
Microbiology	Predictable gut flora	MDR/low-virulence/opportunistic — Enterococcus (incl. VRE), Candida, Pseudomonas aeruginosa, Stenotrophomonas maltophilia, coag-neg staph, MRSA, Acinetobacter
Cultures	Often polymicrobial with predictable gut organisms	Often multiple isolates of hospital-acquired MDR organisms; may be culture-negative despite florid sepsis
Fever pattern	High-grade, swinging	Afebrile or low-grade — host too sick to mount fever; hypothermia a poor sign
WCC	Leucocytosis with left shift	Often leucopenia or normal WCC in immunoparalysed host
Treatment	Surgery (source control) + narrow antibiotics	Broad-spectrum anti-MDR + antifungal + immune-supportive care; surgery often unhelpful (no source to fix)
Mortality	5-20%	30-50% — driven by multi-organ failure rather than uncontrolled infection per se
Conceptual model	A surgical disease	A marker of critical illness / immune dysfunction — "the peritoneum as the organ that fails"[5]

Organisms in tertiary peritonitis and rationale for cover

Organism	Why it dominates in tertiary peritonitis	Antibiotic of choice
Enterococcus (E. faecium → VRE, E. faecalis)	Selected by cephalosporin use (inherently resistant); intrinsically resistant gut coloniser	Ampicillin (E. faecalis); linezolid or daptomycin for VRE (E. faecium)
Candida (albicans, glabrata, krusei)	Broad-spectrum antibiotics deplete bacteria; TPN; central lines; multiple laparotomies	Echinocandin (caspofungin) initially; fluconazole step-down
Pseudomonas aeruginosa	Biofilm on catheters/drains; prior antibiotic exposure selects MDR strains	Piperacillin/tazobactam, ceftolozane/tazobactam, ceftazidime-avibactam, or meropenem (depending on sensitivities); add aminoglycoside in severe
Stenotrophomonas maltophilia	Selected by carbapenems; environmental water organism in ICU sinks	TMP-sulfamethoxazole (only reliable agent)
MRSA	Lines, wounds, prolonged ICU stay	Vancomycin or linezolid
Acinetobacter baumannii	survives on ICU surfaces; extreme MDR	Sulbactam combinations, polymyxin/colistin, tigecycline (organism-dependent)
Coagulase-negative staphylococci	Foreign material (mesh, drains), low-virulence	Often contaminants; treat only if repeatedly isolated + clinical infection

[1]

Approach to tertiary peritonitis — the patient who isn't getting better

Confirm the diagnosis — persistent or recurrent intra-abdominal sepsis ≥48 h after adequate source control of secondary peritonitis. Low-grade fever, persistent leucocytosis/thrombocytopenia, rising lactate, ongoing vasopressor requirement, persistent ileus.
Re-image — CT abdomen/pelvis with contrast looking for undrained collection, leak, ischaemia, abscess. False negatives are common in tertiary peritonitis — small loculated collections may not be drainable.
Re-evaluate source control — is there a drainable collection? A missed enterocutaneous fistula? Anastomotic leak? Consider diagnostic laparoscopy/laparotomy if imaging equivocal — BUT surgery in tertiary peritonitis is often unrewarding ("sticky" peritoneum, no clear source, high surgical risk).
Re-culture aggressively — peritoneal fluid (from drains), blood cultures, line tips. Send for fungal and atypical cultures. Antimicrobial stewardship review.
Broaden empirically — anti-MDR cover: meropenem + vancomycin + echinocandin (caspofungin). Adjust to local antibiogram and isolate sensitivities. Add aminoglycoside or colistin for extreme MDR GNR.
Address the failing host — this is a disease of immune paresis. Enteral nutrition (gut mucosal trophism, reduces translocation), tight glycaemic control (6-10 mmol/L), lung-protective ventilation, renal support, avoid iatrogenic immunosuppression where possible. Consider IVIG/G-CSF only in selected cases (no routine evidence).
Reassess daily — serial CRP/procalcitonin trends, lactate clearance, vasopressor weaning, organ support weaning. De-escalate antibiotics as soon as sensitivities allow.
Mortality is 30-50% — discuss goals of care early with family and MDT. Persistent multi-organ failure beyond day 7-10 of tertiary peritonitis carries >80% mortality.[5]

[1]

PD peritonitis — organisms, source and expected course

Organism	% of cases	Source	Typical course
Coagulase-negative staphylococci (S. epidermidis)	30-45%	Touch contamination at exchange — most common, usually mild	Responds to intraperitoneal (IP) vancomycin/cefazolin; catheter usually retained
Staphylococcus aureus	10-20%	Exit-site / tunnel infection; biofilm on catheter	More severe; may need catheter removal if tunnel infection or refractory
Streptococcus / Enterococcus	10-15%	Oropharyngeal/gut source; trans-visceral migration	Treat with IP vancomycin/cefazolin + IP aminoglycoside; Enterococcus — ampicillin if susceptible
Gram-negative bacilli (E. coli, Klebsiella, Pseudomonas)	15-25%	Gut translocation, urinary source; Pseudomonas from water/biofilm	Pseudomonas → catheter removal usually required; cover with IP ceftazidime/gentamicin, then per sensitivities
Fungal (Candida)	2-5%	Heavily antibiotic-exposed, immunosuppressed, recent bacterial peritonitis	Catheter removal MANDATORY; systemic antifungal (echinocandin/fluconazole); mortality 10-25%
Culture-negative	10-20%	—	Usually responds to IP antibiotics; rule out fastidious organisms (mycobacteria, fungi) if persistent

[1]

Indications for Tenckhoff catheter removal in PD peritonitis

Indication	Reason
Fungal peritonitis (any Candida)	Catheter universally infected/biofilm-coated; cure impossible with catheter in situ. Remove immediately on diagnosis.
Refractory peritonitis (no clinical/culture response at 5 days of appropriate IP antibiotics)	Biofilm; needs catheter removal + 2-3 weeks of systemic antibiotics before reinsertion
Recurrent peritonitis (same organism <30 days after completion of therapy)	Suggests biofilm or tunnel focus; remove catheter
Pseudomonas / Stenotrophomonas	Often refractory; catheter removal usually required to cure
Repeating peritonitis (different organism <30 days after therapy)	Often dialysate/tubing contamination; review exchange technique
Tunnel infection not responding to antibiotics	Suggests catheter focus; removal prevents recurrent
Severe intra-abdominal pathology (appendicitis, ischaemia, perforation)	Differentiate from PD peritonitis (PD = monomicrobial usually); treat as secondary peritonitis + remove catheter

[1]

Empiric management of suspected PD peritonitis — intraperitoneal antibiotics

Diagnosis — cloudy dialysate + abdominal pain ± fever. The cardinal sign is cloudy effluent. Send: dialysate cell count + differential (after ≥2 h dwell), Gram stain, culture. PD peritonitis = dialysate WCC >100/mm³ with PMN >50% (after a dwell).[6]
Empiric IP antibiotics within 2-3 h of presentation — IP vancomycin or cefazolin (gram-positive cover) + IP ceftazidime or gentamicin (gram-negative cover). IP dosing delivers the antibiotic directly to the peritoneum — higher local levels than IV. Loading dose added to one exchange; dwell ≥6 h.
Add heparin 500-1000 U/L to dialysate — prevents fibrin clots obstructing the catheter.
Once culture results (24-72 h) — narrow to organism. CoNS → IP cefazolin or vancomycin alone; S. aureus → IP vancomycin + review exit site; gram-negative → IP ceftazidime (or gentamicin if not aminoglycoside-sparing); fungal → remove catheter + systemic antifungal.
Duration: 2 weeks (coNS, strep); 3 weeks (S. aureus, gram-negative, Enterococcus); until catheter removed + 2-3 weeks systemic (fungal).
Catheter removal if: fungal, refractory at 5 days, recurrent (same organism), Pseudomonas/Stenotrophomonas, or severe intra-abdominal pathology.
Prevention — meticulous exit-site care (mupirocin daily to nares + exit site), training/re-training on aseptic technique (disconnect systems reduce touch contamination), prophylactic mupirocin/rifampicin for S. aureus nasal carriage.

Routes of antibiotic administration in PD peritonitis — IP vs IV

Route	Pros	Cons	When
Intraperitoneal (IP)	Preferred — achieves high peritoneal levels; outperforms IV for PD peritonitis; avoids systemic exposure	Adds to dialysate; requires dwell ≥6 h; compatibility (vancomycin + ceftazidime may precipitate — give in separate exchanges)	First-line for all PD peritonitis; load in the long (overnight) dwell.[6]
Intravenous (IV)	Easy, well tolerated	Lower peritoneal penetration than IP; less effective	Reserve for severe sepsis, systemic infection, or when PD temporarily held; combine with IP

Additional clinical pearls

SBP — diagnosis, treatment and prevention pearls

Diagnostic paracentesis is the single most important investigation in a cirrhotic with ascites and ANY decompensation. Up to a third of SBP cases are culture-negative — PMN ≥ 250 cells/mm³ is the diagnostic threshold, NOT culture positivity. Treating SBP does not require a positive culture.[1] }
Bedside inoculation of ascites into blood-culture bottles doubles yield to 70-80%. Send 10 mL into each bottle at the bedside — the lab-delayed inoculated swab misses half of all SBP.[1] }
Reagent strip (leukocyte esterase dipstick) PMN surrogate is a fast bedside test when lab turnaround is slow — a ≥2+ reading has ~90% sensitivity for SBP.[1] }
Cirrhotic coagulopathy is NOT a contraindication to paracentesis. Mild bleeding <1%, life-threatening bleeding <0.1%. Do NOT give prophylactic FFP/platelets before a tap (these do not prevent the rare visceral injury bleeds anyway).[1] }
Coagulopathy and thrombocytopenia in cirrhosis reflect rebalanced haemostasis — the INR is a poor predictor of bleeding in cirrhosis (pro- and anti-coagulant factors fall in parallel). DO NOT let an abnormal INR delay a diagnostic tap.[1] }
The 48-hour re-tap is essential — repeat paracentesis. PMN should fall by ≥25% at 48 h on appropriate antibiotic. Failure to fall → resistant organism, missed secondary peritonitis, or inadequate source control. Widen empiric cover and image.[1] }
Five days is as effective as ten for uncomplicated SBP — short course reduces resistance and C. difficile. Extend to 7 days only if sepsis, bacteraemia, or delayed response.[1] }
Albumin works by FOUR mechanisms — oncotic (effective circulating volume), anti-inflammatory (binds TNF/endotoxin), renal perfusion (prevents HRS), and reduces bacterial translocation. Sort 1999: HRS 33%→10%, mortality 29%→10%. Albumin dose: 1.5 g/kg day 1, 1 g/kg day 3.[2] }
SBP prophylaxis indications — (a) Previous SBP (lifelong until transplant — recurrence 70% at 1 year without prophylaxis vs 20% with). (b) GI bleed in cirrhosis (7-day course of norfloxacin/ciprofloxacin + ceftriaxone in severe cirrhosis — the ONLY primary prophylaxis with RCT evidence). (c) Low ascitic protein (<15 g/L) WITH impaired renal/liver function (creatinine ≥106 μmol/L, BUN ≥25, sodium ≤130, Child-Pugh ≥9 + bili ≥22 — primary prophylaxis by norfloxacin). (d) All cirrhotic inpatients with ascites — broad-screen on admission is NOT routine, but a low threshold to start prophylaxis after recovery from a first SBP is.[1] }
Quinolone-resistance is rising — long-term norfloxacin prophylaxis selects resistant gut flora. Up to 50% of breakthrough SBP on norfloxacin is gram-positive or quinolone-resistant gram-negative. NEVER use oral ofloxacin/ciprofloxacin empirically for quinolone-prophylaxis SBP — use cefotaxime/ceftriaxone.[3] }
SBP doubles 1-year mortality (30-50%) even with optimal treatment. SBP is a decompensating event — REFER FOR LIVER TRANSPLANT EVALUATION after the first episode. MELD-Na drives listing.[1] }
Bacteraemia occurs in 30-50% of SBP — send blood cultures with the paracentesis. Bacteraemic SBP has worse outcomes (higher HRS risk, mortality). Albumin is even more important in this group.[2] }

Secondary and tertiary peritonitis pearls

Source control is MANDATORY for secondary peritonitis — antibiotics alone cannot cure it. The intra-abdominal source (perforation, leak, abscess, necrosis) must be addressed by drainage, debridement, repair, or resection. Antibiotic penetration into pus/necrotic tissue is poor.[3] }
Surviving Sepsis hour-1 bundle: lactate, blood cultures, antibiotics, 30 mL/kg crystalloid, vasopressors to MAP ≥65. Source control within 6 h of septic shock — each hour of delay in the hypotensive patient increases mortality by ~7-8%.[11] }
Piperacillin/tazobactam is the workhorse empiric regimen for severe community-acquired or healthcare-associated intra-abdominal infection — covers Enterococcus, ESBL GNR, anaerobes, and most Pseudomonas. Switch to meropenem if prior carbapenem-sparing failure, ESBL colonisation, or severe sepsis post-major surgery.[7] }
Cover Enterococcus in healthcare-associated / postoperative IAI — E. faecium is selected by cephalosporins and is the dominant Enterococcus in nosocomial infection; ~50% are VRE in some ICUs. Add vancomycin empirically in post-op/nosocomial IAIs.[7] }
Add an echinocandin (caspofungin) for Candida peritonitis risk factors — recent fluoroquinolone/cephalosporin exposure, TPN, multiple laparotomies, necrotising pancreatitis, immunosuppression, prolonged ICU stay. Switch to fluconazole if stable and C. albicans susceptible.[10] }
STOP-IT: 4 days of antibiotics after adequate source control — non-inferior to ~8 days. Prolonged courses drive resistance and C. difficile without benefit. Continue beyond 4 days ONLY if: bacteraemia/fungaemia, persistent sepsis, incomplete source control, or immunosuppression.[4] }
Use WATER-SOLUBLE contrast (gastrograffin/omnipaque) — NEVER barium — for suspected anastomotic leak or perforation. Barium in the peritoneum causes severe granulomatous peritonitis and is catastrophic if surgery follows.[11] }
Damage-control surgery for the unstable patient: temporary abdominal closure (VAC / Bogota bag), planned second-look laparotomy in 24-48 h. Definitive anastomosis is deferred until physiology restored — primary anastomosis in an unstable, acidotic, coagulopathic patient will leak.[11] }
Tertiary peritonitis is a disease of the failing host, not uncontrolled infection. Cultures are often MDR/low-virulence organisms (Enterococcus, Candida, Pseudomonas, Stenotrophomonas). Surgery often finds "sticky" peritoneum without a clear source — the operation is unrewarding. Focus on broad anti-MDR cover, antifungal, enteral nutrition, and supportive ICU care.[5] }
Stenotrophomonas in tertiary peritonitis → think carbapenem exposure. The only reliable agent is trimethoprim-sulfamethoxazole. A Stenotrophomonas isolate in abdominal cultures tells you the patient has had prolonged carbapenem exposure and immune paresis.[5] }
Tertiary peritonitis may be culture-NEGATIVE despite florid sepsis — the host is too immunoparalysed to mount a neutrophil response, and antibiotics have sterilised cultures. Persistent leucopenia, thrombocytopenia, rising lactate, vasopressor dependence are markers. Treat empirically with broad-spectrum + antifungal.[5] }
Candida peritonitis is a marker of severe critical illness, not the cause. Mortality 40-60% even with appropriate antifungal therapy. Source control (remove infected material, drain collections) and host support matter as much as the antifungal.[10] }
The immunocompromised host (transplant, chemotherapy, neutropenia, chronic steroids) presents atypically — often afebrile, blunted abdominal signs, late presentation. Lower threshold for CT and surgical source control. WSES 2021 immunocompromised guidelines recommend early imaging and broad empiric cover.[12] }
Intra-abdominal hypertension (IAH) and abdominal compartment syndrome (ACS) complicate severe peritonitis — measure intravesical pressure; IAP >12 mmHg = IAH, >20 mmHg with new organ failure = ACS. Surgical decompression (open abdomen) may be required to relieve ACS, prevent renal failure, and improve ventilation.[11] }
Procalcitonin trends can guide antibiotic duration in intra-abdominal infection — falling levels support early stopping at 4 days. A persistently elevated procalcitonin suggests ongoing infection, undrained source, or tertiary peritonitis.[11] }

Catheter-related (PD) peritonitis and special situations

Cloudy dialysate + abdominal pain = PD peritonitis until proven otherwise. Send the FIRST cloudy bag for cell count + culture. WCC >100/mm³ with PMN >50% (after ≥2 h dwell) confirms PD peritonitis.[6] }
Intraperitoneal (IP) antibiotics outperform IV in PD peritonitis. IP vancomycin/cefazolin (gram-positive) + IP ceftazidime/gentamicin (gram-negative) as initial empiric therapy, loaded in the long (overnight) dwell. Vancomycin + ceftazidime may precipitate — give in separate exchanges.[6] }
Add heparin 500-1000 U/L to dialysate — prevents fibrin clots that obstruct the catheter and reduce antibiotic delivery during peritonitis. }
Coagulase-negative staphylococcal PD peritonitis = touch contamination. Mild course, usually responds to IP cefazolin/vancomycin, catheter retained. Review exchange technique (the patient re-touches the disconnect system with imperfect asepsis).[6] }
Staphylococcus aureus PD peritonitis — examine the exit site and tunnel. S. aureus often comes from an exit-site/tunnel infection; cure is unlikely without addressing the catheter focus. Nasal mupirocin reduces S. aureus carriage and exit-site infection.[6] }
Pseudomonas / Stenotrophomonas PD peritonitis usually needs catheter removal. Biofilm-coated catheters are impossible to sterilise with antibiotics alone. Combine IP ceftazidime/gentamicin with catheter removal + 2-3 weeks of systemic antibiotics.[6] }
Fungal PD peritonitis → remove the catheter IMMEDIATELY on diagnosis. Mortality 10-25%; cure impossible with catheter in situ. Start systemic echinocandin → fluconazole step-down. Risk factors: recent bacterial peritonitis (prolonged antibiotics), immunosuppression, diabetes.[6] }
Refractory PD peritonitis (no clinical/culture response at 5 days) → catheter removal. Biofilm-coated catheter, infection with Pseudomonas/fungal, or tunnel infection. Continuing IP antibiotics beyond 5 days without response risks systemic sepsis and peritoneal damage.[6] }
PD peritonitis is monomicrobial in 80% of cases — polymicrobial PD peritonitis (especially with anaerobes) suggests an intra-abdominal source (perforation, ischaemia) — treat as secondary peritonitis and image. Catheter removal often required.[6] }
Antibiotic prophylaxis before PD catheter placement — single dose of IV cefazolin at induction reduces early PD peritonitis. Exit-site mupirocin daily reduces S. aureus exit-site infection and peritonitis.[6] }

Additional red flags

Critical peritonitis red flags — the warning features

SBP with rising creatinine or bilirubin >68 μmol/L → give albumin 1.5 g/kg day 1, 1 g/kg day 3 (non-negotiable; prevents HRS, halves mortality).[2] }
SBP not responding at 48 h (PMN fall <25%) → resistant organism or missed secondary peritonitis — broaden to meropenem, repeat the tap, image for perforation.[1] }
Polymicrobial ascitic culture or anaerobes → this is NOT SBP. Suspect perforation → CT + surgery.[1] }
Quinolone-prophylaxis patient with SBP → do NOT use oral ofloxacin/ciprofloxacin. Use cefotaxime/ceftriaxone — high quinolone resistance.[3] }
Tertiary peritonitis → broad-spectrum anti-MDR + echinocandin, address immune paresis, mortality 30-50%.[5] }
Stenotrophomonas in tertiary peritonitis → only reliable agent is trimethoprim-sulfamethoxazole; selected by carbapenem exposure.[5] }
Candida peritonitis in critically ill / postoperative → echinocandin (caspofungin), source control, mortality 40-60%.[10] }
Septic shock from abdominal source → Surviving Sepsis hour-1 bundle; source control within 6 h.[11] }
Anastomotic leak post-surgery (day 3-7) → fever, ileus, drain output (faeculent/bilious) → re-operation; image with water-soluble contrast NEVER barium.[3] }
Abdominal compartment syndrome (intra-abdominal pressure >20 mmHg with new organ failure) → surgical decompression, open abdomen.[11] }
PD peritonitis with fungal / Pseudomonas / refractory at 5 days / recurrent same organism → catheter removal MANDATORY.[6] }
Polymicrobial PD peritonitis (with anaerobes) → intra-abdominal source (perforation), not PD peritonitis — treat as secondary peritonitis.[6] }
Cirrhotic with "valentino's sign" (RIF pain from gastric fluid tracking down the right paracolic gutter) → perforated PUD, erect CXR for free gas.[3] }
Immunosuppressed patient (transplant, chemo, neutropenia) with abdominal pain → atypical, often afebrile; early CT + surgical source control; broad empiric cover per WSES 2021.[12] }

Summary — what the CICM/FFICM/EDIC candidate must know

The 10 facts that come up in the exam

SBP diagnosis = ascitic PMN ≥ 250 cells/mm³ — culture often negative, do not wait.
SBP treatment = cefotaxime/ceftriaxone + ALBUMIN (1.5 g/kg D1, 1 g/kg D3) — albumin halves HRS and mortality.
SBP organisms: monomicrobial — E. coli, Klebsiella, S. pneumoniae (anaerobes rare).
Secondary peritonitis = polymicrobial + surgical source control MANDATORY; antibiotics alone insufficient.
STOP-IT: 4 days of antibiotics after adequate source control (no benefit of longer).
Tertiary peritonitis = MDR organisms (Enterococcus, Candida, Pseudomonas) + 30-50% mortality; host immune failure.
CAPD peritonitis = cloudy dialysate, WCC >100/mm³ with PMN >50%; intraperitoneal antibiotics first-line.
Catheter removal in PD peritonitis: fungal, refractory, recurrent, Pseudomonas.
Surviving Sepsis hour-1 bundle + source control within 6 h for septic shock from abdominal source.
Albumin trials: Sort 1999 (SBP, benefit), ANSWER 2018 (long-term outpatient, benefit), ATTIRE 2021 (inpatient target, no benefit) — be selective.

[1]

Peritonitis at a glance — the one-table revision

Feature	SBP	Secondary	Tertiary	CAPD
Population	Cirrhosis + ascites	Any abdominal catastrophe	Critically ill ICU patient	PD patient
Source	None (haematogenous)	Perforation/leak/abscess	Persistent/recurrent; often no source	Catheter (touch contamination)
Microbiology	Monomicrobial GNR (E. coli, Klebsiella, S. pneumo)	Polymicrobial gut flora + anaerobes	MDR (Enterococcus, Candida, Pseudomonas, MRSA)	Staph (CoNS, S. aureus), GNR, fungal
Diagnosis	Ascitic PMN ≥ 250/mm³	CT + surgery	Persistent sepsis, cultures positive	Dialysate WCC >100/mm³, PMN >50%
First-line treatment	Cefotaxime + ALBUMIN	Surgery + piperacillin/tazobactam	Meropenem + vancomycin + echinocandin	Intraperitoneal vancomycin/cefazolin + ceftazidime
Source control	NOT required (no source)	MANDATORY	Often unrewarding	Catheter removal for specific indications
Antibiotic duration	5-7 days	4 days after source control	Often prolonged (until clinically resolving)	2-3 weeks
Mortality	10-30% (albumin halves high-risk)	5-20% (driven by source-control timing)	30-50%	<5% per episode
Key trial	Sort 1999 (albumin)	STOP-IT 2015 (4 days)	Evans 2001 (tertiary phenotype)	ISPD 2016 (IP antibiotics)

[1]

References

[1]Runyon BA, AASLD. Introduction to the revised American Association for the Study of Liver Diseases Practice Guideline management of adult patients with ascites due to cirrhosis 2012 Hepatology, 2013.PMID 23463403
[2]Sort P, Navasa M, Arroyo V, et al. Effect of intravenous albumin on renal impairment and mortality in patients with cirrhosis and spontaneous bacterial peritonitis N Engl J Med, 1999.PMID 10432325
[3]Solomkin JS, Mazuski JE, Bradley JS, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America Clin Infect Dis, 2010.PMID 20034345
[4]Sawyer RG, Claridge JA, Nathens AB, et al. Trial of short-course antimicrobial therapy for intraabdominal infection N Engl J Med, 2015.PMID 25992746
[5]Evans HL, Raymond DP, Pelletier SJ, et al. Tertiary peritonitis (recurrent diffuse or localized disease) is not an independent predictor of mortality in surgical patients with intraabdominal infection Surg Infect (Larchmt), 2001.PMID 12593701
[6]Li PK-T, Szeto CC, Piraino B, et al. ISPD Peritonitis Recommendations: 2016 Update on Prevention and Treatment Perit Dial Int, 2016.PMID 27282851
[7]Mazuski JE, Tessier JM, May AK, et al. The Surgical Infection Society Revised Guidelines on the Management of Intra-Abdominal Infection Surg Infect (Larchmt), 2017.PMID 28085573
[8]Caraceni P, Riggio O, Angeli P, et al.; ANSWER Study Investigators. Long-term albumin administration in decompensated cirrhosis (ANSWER): an open-label randomised trial Lancet, 2018.PMID 29861076
[9]China L, Freemantle N, Forrest E, et al.; ATTIRE trial investigators. A Randomized Trial of Albumin Infusions in Hospitalized Patients with Cirrhosis N Engl J Med, 2021.PMID 33657293
[10]Pappas PG, Kauffman CA, Andes D, et al. Clinical practice guidelines for the management of candidiasis: 2009 update by the Infectious Diseases Society of America Clin Infect Dis, 2009.PMID 19191635
[11]Sartelli M, Mazuski JE, Coccolini F, et al. Intra-abdominal infections survival guide: a position statement by the Global Alliance For Infections In Surgery World J Emerg Surg, 2024.PMID 38851700
[12]Coccolini F, Sartelli M, Podda M, et al. Acute abdomen in the immunocompromised patient: WSES, SIS-E, WSIS, AAST, and GAIS guidelines World J Emerg Surg, 2021.PMID 34372902

[1]

Types of peritonitis

Feature	SBP	Secondary	Tertiary	CAPD
Source	None (spontaneous)	Perforation, leak, abscess	Persistent/recurrent after secondary	Peritoneal dialysis catheter
Population	Cirrhosis with ascites	Anyone with abdominal pathology	Critically ill ICU patients	PD patients
Organisms	Monomicrobial (E. coli, Klebsiella)	Polymicrobial (gram-negative, anaerobes)	Resistant (Enterococcus, Candida, Pseudomonas, MRSA)	Staphylococcus (epidermidis, aureus)
Diagnosis	Ascitic PMN >250/mm³	CT, laparotomy	Persistent sepsis, cultures positive	Cloudy dialysate, PMN >100/mm³
Treatment	Cefotaxime + ALBUMIN	SURGERY (source control) + broad antibiotics	Anti-fungal, anti-Enterococcus, anti-Pseudomonas	Intraperitoneal antibiotics ± catheter removal
Mortality	10-30%	5-20%	30-50%	<5% (per episode)

[1]

Management of suspected secondary peritonitis

Resuscitate — ABC. IV fluids (septic patients often hypovolaemic). Vasopressors (noradrenaline) for shock. Broad-spectrum antibiotics within 1h
CT abdomen with contrast — identify source (perforation, abscess, leak, ischaemia)
Source control — SURGERY is MANDATORY: (a) Perforation — repair/resect. (b) Abscess — drain (percutaneous or surgical). (c) Leak — repair/re-anastomose/stoma. (d) Ischaemia — resect necrotic bowel. (e) Lavage (wash out peritoneum)
Antibiotics — broad-spectrum, polymicrobial cover: piperacillin/tazobactam OR meropenem + metronidazole. Cover gram-negative, anaerobes, Enterococcus. Consider antifungal (candida) if prolonged, immunosuppressed
ICU support — ventilation, renal support, vasopressors
Duration: 4 days after source control (STOP-IT trial showed no benefit of longer)
Monitor for tertiary peritonitis — persistent sepsis despite source control → resistant organisms → broaden antibiotics

[1] [1]

SAQ — Spontaneous bacterial peritonitis in a cirrhotic with decompensation

10 minutes · 10 marks

[1]

SAQ — Secondary peritonitis from a perforated viscus with septic shock

10 minutes · 10 marks

[1]

Clinical pearls

High-yield peritonitis points for CICM/FFICM exam

SBP = ascitic PMN >250 cells/mm³. Diagnostic paracentesis is MANDATORY in any cirrhotic with ascites and new fever/abdominal pain/altered mental state/renal failure. DO NOT wait for culture (50% culture-negative) — treat based on PMN count.[1] }
SBP organisms are usually MONOMICROBIAL (unlike secondary peritonitis which is polymicrobial). Gram-negative rods (E. coli 70%, Klebsiella) most common. Gram-positive (Streptococcus pneumoniae, Enterococcus) in 25%. Anaerobes RARE (protected by low ascitic oxygen).[1] }
Albumin in SBP — reduces hepatorenal syndrome and mortality. Landmark trial (Sort 1999, NEJM): SBP patients given cefotaxime + albumin vs cefotaxime alone. Albumin REDUCED HRS (10% vs 33%) and mortality (10% vs 29%). Dose: 1.5 g/kg day 1, 1 g/kg day 3. Give to ALL SBP patients (especially with risk factors: bilirubin >68, creatinine >88).[2] }
SBP treatment: cefotaxime 2g IV BD-TDS for 5-7 days. Alternative: ceftriaxone 2g IV OD. Oral ofloxacin (400mg BD) for uncomplicated SBP if not septic and no prior antibiotic prophylaxis. 5 days is as effective as 10 days.[1] }
SBP prophylaxis: norfloxacin 400mg OD or ciprofloxacin 500mg OD. Indications: (1) Previous SBP (lifelong prophylaxis). (2) Ascitic protein <15 g/L (especially if also renal/liver dysfunction). (3) GI bleed in cirrhotic (prevents SBP and bacterial translocation). Primary prophylaxis controversial.[1] }
Secondary peritonitis: source control is MANDATORY. Antibiotics alone CANNOT cure secondary peritonitis — the intra-abdominal source (perforation, leak, abscess) must be addressed surgically. Delay in source control >24h increases mortality significantly.[3] }
Secondary peritonitis is POLYMICROBIAL. Gram-negative (E. coli, Klebsiella), anaerobes (Bacteroides, Clostridium), Enterococcus. Antibiotics: piperacillin/tazobactam OR meropenem + metronidazole. Cover Enterococcus (especially post-operative, healthcare-associated).[3] }
STOP-IT trial (2015): 4 days antibiotics after source control. RCT: complicated intra-abdominal infection. 4 days vs ~8 days after source control. Result: NO difference in outcomes (infection-related adverse events, mortality). SHORT course (4 days after adequate source control) is sufficient.[4] }
Tertiary peritonitis: persistent/recurrent after treatment of secondary. Occurs 30-50% of patients after secondary peritonitis. Organisms: RESISTANT — Enterococcus (VRE), Candida, Pseudomonas, MRSA, Stenotrophomonas. Reflects immune dysfunction of critically ill. Mortality 30-50%. Treatment: broad-spectrum (vancomycin, meropenem, echinocandin/fluconazole), anti-fungal essential.[5] }
Candida peritonitis. Common in: post-operative, prolonged antibiotics, immunosuppressed, multiple abdominal surgeries, severe pancreatitis. Treatment: fluconazole (if susceptible) or echinocandin (caspofungin, micafungin) if critically ill/unstable. Remove infected foreign material. Source control essential.[3] }
CAPD peritonitis: cloudy dialysate + abdominal pain. Diagnosis: dialysate WBC >100/mm³ (after dwell), PMN >50%. Organisms: coagulase-negative Staph (40% — from touch contamination), Staph aureus (15% — exit-site infection), gram-negative (15%), fungal (2% — needs catheter removal). Treatment: INTRAPERITONEAL antibiotics (vancomycin/cefazolin + ceftazidime/gentamicin). Catheter removal if: fungal, refractory, recurrent, Pseudomonas.[6] }
Post-operative peritonitis (anastomotic leak). Presents day 3-7 post-surgery with fever, ileus, abdominal pain, sepsis, drain output (faeculent/bilious). CT with water-soluble contrast (water-soluble — safer than barium if leak). Management: re-operation (repair/defunction), antibiotics, ICU.[3] }
Abdominal sepsis is a common cause of septic shock in ICU. Always consider intra-abdominal source in septic patient without obvious cause. CT abdomen is essential. Source control (surgery/drainage) is as important as antibiotics. Procalcitonin may help (elevated in bacterial infection).[3] }
Diagnostic peritoneal lavage (DPL) — largely replaced by FAST ultrasound and CT. Still used occasionally in trauma (if FAST negative but suspicion high). Positive: >100,000 RBC/mL (indicating hollow viscus injury requiring laparotomy) or bile/food/faeces (bowel injury).[3] }

Red flags

Prognosis

Albumin in SBP (Sort 1999, NEJM)

RCT: 126 cirrhotic patients with SBP. Cefotaxime + albumin vs cefotaxime alone.

Hepatorenal syndrome: 10% (albumin) vs 33% (control) — p=0.002, 67% relative reduction
In-hospital mortality: 10% (albumin) vs 29% (control) — p=0.01, 65% relative reduction
3-month mortality: 22% vs 41% — p=0.03 [1]

Albumin dose: 1.5 g/kg day 1, 1 g/kg day 3. CONCLUSION: Albumin is STANDARD OF CARE for SBP (especially with bilirubin >68 μmol/L or creatinine >88 μmol/L).

SBP — spontaneous bacterial peritonitis: deep dive [1]

Pathophysiology and portal of entry — why the four peritonitides differ

Type	Source / portal of entry	Microbiology	Key host factor
SBP	Haematogenous / lymphatic seeding of ascites; intestinal bacterial translocation across oedematous, more-permeable cirrhotic gut wall. NO visible intra-abdominal source on imaging.	Monomicrobial — single organism in 90%+ of culture-positive cases	Cirrhosis with ascites: low ascitic opsonic/complement activity (protein <15 g/L, C3 low) impairs phagocytosis. Diseased Kupffer cells fail to clear portal bacteraemia.
Secondary	Spillage from a breached viscus — perforation, ischaemic necrosis, anastomotic leak, abscess, trauma. Source is anatomical and IMAGING-DETECTABLE.	Polymicrobial — gut flora: gram-negative bacilli, anaerobes (Bacteroides, Clostridium), Enterococcus, streptococci; ± Candida	Any patient with abdominal pathology; severity depends on load, virulence, host, and source-control TIMING.
Tertiary	Persistent or recurrent infection after apparent control of secondary peritonitis — failure of source control or host immunity. Often NO clear new source.	MDR/fermenters/fungi — Enterococcus (incl. VRE), Candida, Pseudomonas, Stenotrophomonas, MRSA, Acinetobacter	Critically ill ICU patient: immune paralysis, prolonged antibiotics, foreign material (mesh, drains), open abdomen, TPN. Reflects immunoparesis more than new contamination.
CAPD / catheter-related	Inoculation along the Tenckhoff catheter (touch contamination, exit-site/tunnel infection) or trans-visceral migration. Peritoneum is the dialyser.	Staphylococci (CoNS 30-45%, S. aureus 10-20%), gram-negatives (15-25%), Streptococcus (10%), fungal (2-5%)	Peritoneal dialysis patients; defence diluted by ~2 L dialysate, low opsonin, catheter as foreign body.

[1]

Ascitic fluid analysis — interpreting the diagnostic paracentesis

Result	Definition	Significance	Action
SBP	Ascitic PMN ≥ 250 cells/mm³ (with or without positive culture)	Diagnostic of SBP — treat empirically NOW. Culture positive in only ~50%.	Cefotaxime 2 g IV BD-TDS × 5-7 days + albumin 1.5 g/kg day 1, 1 g/kg day 3.[1]
Culture-negative neutrocytic ascites (CNNA)	PMN ≥ 250/mm³ AND negative culture (and no antibiotics in prior 7 days)	A variant of SBP — identical prognosis and management. The 50% culture-negative group.	Treat as SBP. Not a false positive — usually due to culture technique; bedside inoculation into blood-culture bottles ↑ yield to 70-80%.[1]
Monomicrobial non-neutrocytic bacterascites (MNB)	Single organism on culture AND PMN < 250/mm³	Often transient/colonisation that self-clears; BUT treat if symptomatic (fever, pain) or cirrhotic — risk of progression to SBP.	Treat if symptomatic; repeat tap in 24-48 h if asymptomatic.
Polymicrobial bacterascites	Multiple organisms AND PMN < 250/mm³	Suggests PERFORATION — bowel leak inoculating ascites. Diagnostic clue that this is NOT SBP.	Image (CT) for perforation → surgical source control.
Secondary peritonitis (in cirrhotic with ascites)	PMN ≥ 250 AND polymicrobial culture (multiple organisms, including anaerobes/yeast) AND ascitic: total protein >10 g/L, glucose <2.8 mmol/L, LDH >225 U/L	Differentiate from SBP: SBP is monomicrobial, no anaerobes, normal glucose/LDH. Secondary needs surgery.	CT + surgery; broaden to cover anaerobes + Enterococcus.

Diagnostic workup of a cirrhotic patient with ascites and fever/altered mental state

Diagnostic paracentesis is MANDATORY — every cirrhotic with ascites + any of: fever/abdominal pain/altered mental state/renal impairment/GI bleed/leucocytosis. Also tap on admission and at first sign of decompensation. Up to 30% of asymptomatic cirrhotic inpatients have SBP on admission.[1]
Bedside technique — supine, LLQ or RLQ (midline 2 cm below umbilicus in obese). Clean skin. 20-30 mL aspirated. Inoculate blood-culture bottles directly at the bedside (aerobic + anaerobic) — doubles culture yield vs delayed lab inoculation. Send: cell count + differential, protein, albumin, glucose, LDH, Gram stain, culture.
Reagent strip (dipstick) PMN — a leukocyte esterase dipstick (≥2+) gives a rapid bedside PMN surrogate when the lab count is delayed. Sensitivity for SBP ~90%.
Diagnosis = PMN ≥ 250 cells/mm³ — start antibiotics immediately; do NOT wait for culture. If PMN borderline (e.g., 200-250), repeat in 24-48 h.
Coagulopathy is NOT a contraindication — cirrhotic coagulopathy/ascites do not require fresh frozen plasma/platelets before a therapeutic or diagnostic tap (mild bleeding risk <1%, life-threatening <0.1%).[1]
If PMN ≥ 250 → CT abdomen/pelvis if any feature of secondary peritonitis (polymicrobial Gram stain, anaerobes, ascitic glucose <2.8 or LDH >225, free gas on CXR/CT) to exclude a perforated viscus.
Reassess at 48 h — repeat tap. PMN should fall by ≥ 25% at 48 h with appropriate antibiotic. If not, suspect resistant organism or secondary peritonitis → broaden + image.

Organisms in SBP — predicted microbiology

Group	Frequency	Organisms	Antibiotic implication
Gram-negative bacilli	70-75%	Escherichia coli (most common, ~50%), Klebsiella pneumoniae, other Enterobacterales	Third-generation cephalosporin (cefotaxime/ceftriaxone) — covers ~95% of community-acquired SBP.
Gram-positive cocci	20-25%	Streptococcus pneumoniae (pneumococcus), other streptococci, Enterococcus (more common nosocomial)	Cefotaxime covers streptococci/pneumococcus; Enterococcus needs ampicillin/vancomycin (cover if nosocomial).
Anaerobes	<5% (RARE)	Bacteroides, Clostridium	The low ascitic oxygen tension + presence of complement selects against anaerobes — do NOT add metronidazole routinely for classic SBP.
Nosocomial / healthcare-associated SBP	rising	MDR gram-negatives (ESBL, KPC), Enterococcus (incl. VRE), Pseudomonas	Switch to piperacillin/tazobactam, meropenem, or ceftolozane/tazobactam based on local antibiogram. Cefotaxime fails in ~20% of nosocomial SBP.[3]
Fungal	<2%	Candida	Only in heavily antibiotic-exposed/immunosuppressed — add echinocandin.

SBP antibiotic therapy — empiric regimens by clinical context

Clinical context	First-line	Alternative / allergy	Duration
Community-acquired SBP, uncomplicated	Cefotaxime 2 g IV BD-TDS OR ceftriaxone 2 g IV OD	Levofloxacin 500 mg IV/PO OD or moxifloxacin (if no prior quinolone prophylaxis, no vomiting/shock)	5 days (as effective as 10 days)
SBP with sepsis/septic shock	Cefotaxime/ceftriaxone ± piperacillin/tazobactam (broaden for healthcare acquisition)	Meropenem 1 g IV TDS	7 days; reassess at 48 h by PMN fall
Healthcare-associated / nosocomial SBP	Piperacillin/tazobactam 4.5 g IV TDS-QDS OR meropenem	Add vancomycin if VRE/Enterococcus risk; consider ceftolozane/tazobactam for MDR Pseudomonas	7 days, de-escalate on culture
Quinolone prophylaxis failure	NEVER use ofloxacin/ciprofloxacin empirically — high quinolone resistance	Cefotaxime/ceftriaxone (1st)	5-7 days
Add ALBUMIN to ALL SBP	1.5 g/kg IV day 1, 1 g/kg day 3	(Saline is NOT a substitute — must be albumin)	Days 1 and 3 only

[1]

SBP management algorithm — bedside sequence

Confirm diagnosis — ascitic PMN ≥ 250 cells/mm³.
Cefotaxime 2 g IV BD-TDS (or ceftriaxone 2 g IV OD) — start within 1 h of tap. Send blood cultures (bacteraemia in 30-50%).
Albumin 1.5 g/kg IV on day 1, 1 g/kg IV on day 3 — give to ALL SBP patients; MANDATORY if any of: creatinine >88 μmol/L, BUN >30 mg/dL, bilirubin >68 μmol/L.[2]
Re-tap at 48 h — PMN must fall ≥ 25%. If not: suspect resistant organism or secondary peritonitis → broaden (piperacillin/tazobactam or meropenem), CT to exclude perforation.
Search for and treat precipitant — GI bleed (give SBP prophylaxis), urinary/biliary infection, prior paracentesis without albumin, dehydration, large-volume paracentesis without albumin replacement.
Continue therapy 5 days total for uncomplicated SBP; 7 days for sepsis/bacteraemia.
Start SECONDARY PROPHYLAXIS at discharge — norfloxacin 400 mg OD or ciprofloxacin 500 mg OD (lifelong, until transplant). Recurrence 70% at 1 year without prophylaxis vs 20% with.[1]
Refer for liver transplant evaluation — SBP is a decompensating event that doubles 1-year mortality (30-50%). MELD-Na drives listing priority.

Albumin in cirrhosis — three landmark trials, three different conclusions

Trial	Population	Intervention	Key result	Interpretation
Sort 1999 (NEJM)	126 SBP patients	Cefotaxime + albumin (1.5 g/kg D1, 1 g/kg D3) vs cefotaxime alone	HRS 10% vs 33%; hospital mortality 10% vs 29%	Albumin is STANDARD OF CARE for SBP.[2]
ANSWER 2018 (Lancet)	440 decompensated cirrhosis (no SBP)	Long-term weekly albumin × 18 months vs standard care	50% reduction in SBP, fewer decompensation events, 18-month survival improved (74% vs 62%)	Long-term albumin reduces SBP incidence and improves survival in selected outpatients.[8]
ATTIRE 2021 (NEJM)	777 hospitalised cirrhosis patients	Targeted 20% albumin infusions to keep serum albumin >30 g/L vs standard care	No reduction in new infection, renal dysfunction, or mortality	Albumin infusions to a target level in unselected inpatients do NOT prevent infection. Be selective — benefit proven in SBP and ANSWER-selected outpatients only.[9]

ANSWER trial — Caraceni 2018, Lancet (long-term albumin in decompensated cirrhosis)

Open-label RCT, Italy. 440 patients with decompensated cirrhosis, no SBP at entry.

Intervention: human albumin 1.5 g/kg/week for 2 weeks, then 40 g/week for 18 months, vs standard care.
Primary endpoint — 18-month mortality: 77/220 (albumin) vs 99/220 (standard) — HR 0.66, borderline.
Incident SBP: 10% vs 24% — hazard ratio 0.40 (60% reduction).
Grade 3-4 ascites, renal dysfunction, bacterial infection: all significantly reduced.
CONCLUSION: In selected outpatients with decompensated cirrhosis, long-term albumin reduces SBP incidence and may improve survival. Selective use, not blanket ATTIRE-style infusions.[8]

ATTIRE — China 2021, NEJM (albumin in hospitalised cirrhosis)

Multicentre RCT, UK. 777 hospitalised patients with cirrhosis and serum albumin <30 g/L.

Intervention: 20% human albumin to maintain serum albumin ≥30 g/L (target-led) up to 14 days vs standard care.
Primary composite (new infection, renal dysfunction, death) at 15 days: 29.7% vs 30.2% — no difference.
Adverse events: pulmonary oedema more common with albumin.
CONCLUSION: Routine albumin infusions in unselected hospitalised cirrhosis do NOT improve outcomes. Albumin should be reserved for SBP, large-volume paracentesis (≥5 L), HRS, and selected ANSWER-style outpatients.[9]

Secondary peritonitis: deep dive — sources, organisms, source control

Sources of secondary peritonitis — predicted by anatomical site of perforation

Source	Classic organism profile	Clinical clues	Imaging
Perforated appendicitis	E. coli, Bacteroides fragilis, Streptococcus anginosus group	Migration of periumbilical→RLQ pain; McBurney's point tenderness; young male; perforation → diffuse peritonitis/sepsis	CT: peri-appendiceal fat stranding, free fluid ± free gas, appendicolith
Perforated diverticulitis (Hinchey III-IV)	E. coli, Klebsiella, Bacteroides, Clostridium; ± Enterococcus	LLQ pain in older patient; acute severe abdominal pain + rigid abdomen in Hinchey IV; faeculent peritonitis	CT: sigmoid diverticula, free gas ± faeces; Hinchey stage
Perforated peptic ulcer (PUD)	Streptococcus, E. coli, anaerobes (after several hours)	Sudden "tearing" epigastric pain → rigid abdomen; "Valentino's sign" (RIF pain from gastric fluid tracking down right paracolic gutter); standing CXR: free subdiaphragmatic gas	Erect CXR; CT abdomen; "Valsalva/computed tomography" shows anterior free gas
Ischaemic bowel / mesenteric infarction	E. coli, Klebsiella, Clostridium perfringens, Enterococcus; high risk of bacteraemia	AF, heart failure, vasopressors; "pain out of proportion"; metabolic acidosis, rising lactate; bloody diarrhoea late	CT: pneumatosis intestinalis, portal venous gas, mesenteric vessel occlusion, thickened oedematous loops
Anastomotic leak (postoperative, day 3-7)	Enterococcus (ampicillin-resistant), Candida, Pseudomonas, MDR GNR — reflects hospital flora	Fever, ileus, drain output (faeculent/bilious/purulent), sepsis after GI surgery	CT with water-soluble oral/rectal contrast (NOT barium); contrast extravasation
Biliary source (cholecystitis, cholangitis, bile leak)	E. coli, Klebsiella, Enterococcus, Enterobacter; less often anaerobes (Clostridium perfringens — emphysematous)	RUQ pain, jaundice, Charcot's/Reynold's pentad; post-cholecystectomy/ERCP; bile-stained peritoneal fluid	MRCP/ERCP; CT: wall thickening, pericholecystic fluid, ductal dilatation
Post-traumatic (blunt or penetrating bowel injury)	E. coli, Bacteroides, Streptococcus	Mechanism (seatbelt sign, handlebar, GSW); delayed presentation; persistent tachycardia/lactataemia after trauma	CT: free gas, free fluid, bowel wall thickening, mesenteric stranding

[1]

Predicted organisms and antibiotic cover by source (community vs healthcare-associated)

Severity / context	Predicted cover required	Empiric regimen
Community-acquired, mild-moderate	Gram-negative rods + anaerobes	Ceftriaxone 2 g IV OD + metronidazole 500 mg IV TDS, OR co-amoxiclav 1.2 g IV TDS + metronidazole, OR moxifloxacin 400 mg IV OD (monotherapy)
Community-acquired, high-severity / sepsis	Gram-negative rods + anaerobes + Enterococcus	Piperacillin/tazobactam 4.5 g IV TDS-QDS, OR meropenem 1 g IV TDS + metronidazole (if ESBL risk)
Healthcare-associated / postoperative	Gram-negatives (incl. ESBL/Pseudomonas) + Enterococcus + Candida	Piperacillin/tazobactam OR meropenem + vancomycin (Enterococcus/MRSA) + echinocandin (caspofungin 70 mg then 50 mg OD) if recent fluoroquinolone/cephalosporin exposure, TPN, prolonged ICU stay
Biliary source	Gram-neg + Enterococcus (anaerobes RARE in pure biliary disease unless emphysematous)	Ceftriaxone + metronidazole; piperacillin/tazobactam monotherapy if severe
Candida peritonitis (post-op, multiple abdominal re-explorations, necrotising pancreatitis, immunosuppressed)	C. albicans, C. glabrata, C. krusei	Echinocandin (caspofungin/micafungin/anidulafungin) if critically ill; step down to fluconazole if stable and C. albicans susceptible. Remove infected foreign material; source control.[10]

Principles of SOURCE CONTROL — the four D's

DRAINAGE — percutaneous (image-guided) or surgical drainage of collections/abscess. Aim: evacuating pus and contaminated fluid.
DEBRIDEMENT — excision of necrotic, ischaemic or grossly contaminated tissue (non-viable bowel, omental slough). Necrotic tissue is a culture medium that antibiotics cannot penetrate.
DECOMPRESSION / DIVERSION — relieve obstruction (NGT, rectal tube, defunctioning stoma); prevent ongoing contamination.
DEFINITIVE REPAIR — close or resect the breach: primary repair, resection + anastomosis, resection + stoma (Hartmann's), or controlled fistula formation (open abdomen with VAC, "planned re-laparotomy"). [1]

Management of suspected perforated viscus with septic shock

ABCDE resuscitation — oxygen, two large-bore IVs. Crystalloid 30 mL/kg bolus for sepsis-induced hypoperfusion. Noradrenaline first-line vasopressor if MAP <65 mmHg after fluids. Add vasopressin / adrenaline if refractory.
Broad-spectrum antibiotics within 1 h (sepsis hour-1 bundle) — piperacillin/tazobactam 4.5 g IV ± vancomycin (Enterococcus/MRSA) ± echinocandin if Candida risk. Add metronidazole if using ceftriaxone/cefepime alone.
Stat surgical review + anaesthetics — declare a surgical emergency; consent for laparotomy.
Cross-match 4-6 units, group-and-save, check coagulation, platelets, lactate, venous gas, lactate trend.
CT abdomen/pelvis with IV + water-soluble oral/rectal contrast — only if the patient can be stabilised (CT is inadvisable in extremis — go straight to theatre). Confirm free gas, perforation site, abscess.
Urinary catheter + arterial line + CVC — guide fluid responsiveness and vasopressors; target MAP ≥65, lactate clearance ≥10%/h, ScvO₂ ≥70%.
Theatre — emergency laparotomy: peritoneal lavage, identify and repair/resect source, drain. Damage-control surgery in unstable patients (leave abdomen open, VAC, planned second-look).
Post-op ICU — vasopressor wean, lung-protective ventilation, renal support, glycaemic control 6-10 mmol/L, deep vein thrombosis prophylaxis once bleeding controlled, stress ulcer prophylaxis, enteral nutrition within 24-48 h.
Antibiotic duration: 4 days after adequate source control (STOP-IT).[4] Prolong if persistent sepsis, bacteraemia, fungaemia, source control incomplete.

STOP-IT trial — Sawyer 2015, NEJM (duration of antibiotics after source control)

Multicentre RCT, USA. 518 patients with complicated intra-abdominal infection and adequate source control.

Intervention: antibiotics stopped after a fixed ~4-day course (approx. 96 h) vs continued until physician-satisfied (~8 days).
Primary composite (surgical-site infection, recurrent IAI, death) at 30 days: 21.8% (short) vs 22.3% (long) — non-inferior.
Median duration: 4 days (short) vs 8 days (long).
CONCLUSION: Once adequate source control is achieved, ~4 days of antibiotics is sufficient. Prolonged courses do not improve outcomes and select for resistance/C. difficile.[4]

Tertiary peritonitis: deep dive — the failing host

Secondary vs tertiary peritonitis — distinguishing the ICU phenotype

Feature	Secondary peritonitis	Tertiary peritonitis
Timing	Acute — first presentation of perforation/leak	Persistent or recurrent after ≥48 h of treatment of secondary, OR no source identifiable
Host	Often previously well; new intra-abdominal catastrophe	Critically ill; prolonged ICU stay, multi-organ failure, immunoparesis
Source	Identifiable perforation/leak/abscess	Often NO source found at re-operation; widespread serosal inflammation, "sticky" fibrinous peritoneum
Microbiology	Predictable gut flora	MDR/low-virulence/opportunistic — Enterococcus (incl. VRE), Candida, Pseudomonas aeruginosa, Stenotrophomonas maltophilia, coag-neg staph, MRSA, Acinetobacter
Cultures	Often polymicrobial with predictable gut organisms	Often multiple isolates of hospital-acquired MDR organisms; may be culture-negative despite florid sepsis
Fever pattern	High-grade, swinging	Afebrile or low-grade — host too sick to mount fever; hypothermia a poor sign
WCC	Leucocytosis with left shift	Often leucopenia or normal WCC in immunoparalysed host
Treatment	Surgery (source control) + narrow antibiotics	Broad-spectrum anti-MDR + antifungal + immune-supportive care; surgery often unhelpful (no source to fix)
Mortality	5-20%	30-50% — driven by multi-organ failure rather than uncontrolled infection per se
Conceptual model	A surgical disease	A marker of critical illness / immune dysfunction — "the peritoneum as the organ that fails"[5]

Organisms in tertiary peritonitis and rationale for cover

Organism	Why it dominates in tertiary peritonitis	Antibiotic of choice
Enterococcus (E. faecium → VRE, E. faecalis)	Selected by cephalosporin use (inherently resistant); intrinsically resistant gut coloniser	Ampicillin (E. faecalis); linezolid or daptomycin for VRE (E. faecium)
Candida (albicans, glabrata, krusei)	Broad-spectrum antibiotics deplete bacteria; TPN; central lines; multiple laparotomies	Echinocandin (caspofungin) initially; fluconazole step-down
Pseudomonas aeruginosa	Biofilm on catheters/drains; prior antibiotic exposure selects MDR strains	Piperacillin/tazobactam, ceftolozane/tazobactam, ceftazidime-avibactam, or meropenem (depending on sensitivities); add aminoglycoside in severe
Stenotrophomonas maltophilia	Selected by carbapenems; environmental water organism in ICU sinks	TMP-sulfamethoxazole (only reliable agent)
MRSA	Lines, wounds, prolonged ICU stay	Vancomycin or linezolid
Acinetobacter baumannii	survives on ICU surfaces; extreme MDR	Sulbactam combinations, polymyxin/colistin, tigecycline (organism-dependent)
Coagulase-negative staphylococci	Foreign material (mesh, drains), low-virulence	Often contaminants; treat only if repeatedly isolated + clinical infection

[1]

Approach to tertiary peritonitis — the patient who isn't getting better

Confirm the diagnosis — persistent or recurrent intra-abdominal sepsis ≥48 h after adequate source control of secondary peritonitis. Low-grade fever, persistent leucocytosis/thrombocytopenia, rising lactate, ongoing vasopressor requirement, persistent ileus.
Re-image — CT abdomen/pelvis with contrast looking for undrained collection, leak, ischaemia, abscess. False negatives are common in tertiary peritonitis — small loculated collections may not be drainable.
Re-evaluate source control — is there a drainable collection? A missed enterocutaneous fistula? Anastomotic leak? Consider diagnostic laparoscopy/laparotomy if imaging equivocal — BUT surgery in tertiary peritonitis is often unrewarding ("sticky" peritoneum, no clear source, high surgical risk).
Re-culture aggressively — peritoneal fluid (from drains), blood cultures, line tips. Send for fungal and atypical cultures. Antimicrobial stewardship review.
Broaden empirically — anti-MDR cover: meropenem + vancomycin + echinocandin (caspofungin). Adjust to local antibiogram and isolate sensitivities. Add aminoglycoside or colistin for extreme MDR GNR.
Address the failing host — this is a disease of immune paresis. Enteral nutrition (gut mucosal trophism, reduces translocation), tight glycaemic control (6-10 mmol/L), lung-protective ventilation, renal support, avoid iatrogenic immunosuppression where possible. Consider IVIG/G-CSF only in selected cases (no routine evidence).
Reassess daily — serial CRP/procalcitonin trends, lactate clearance, vasopressor weaning, organ support weaning. De-escalate antibiotics as soon as sensitivities allow.
Mortality is 30-50% — discuss goals of care early with family and MDT. Persistent multi-organ failure beyond day 7-10 of tertiary peritonitis carries >80% mortality.[5]

[1]

PD peritonitis — organisms, source and expected course

Organism	% of cases	Source	Typical course
Coagulase-negative staphylococci (S. epidermidis)	30-45%	Touch contamination at exchange — most common, usually mild	Responds to intraperitoneal (IP) vancomycin/cefazolin; catheter usually retained
Staphylococcus aureus	10-20%	Exit-site / tunnel infection; biofilm on catheter	More severe; may need catheter removal if tunnel infection or refractory
Streptococcus / Enterococcus	10-15%	Oropharyngeal/gut source; trans-visceral migration	Treat with IP vancomycin/cefazolin + IP aminoglycoside; Enterococcus — ampicillin if susceptible
Gram-negative bacilli (E. coli, Klebsiella, Pseudomonas)	15-25%	Gut translocation, urinary source; Pseudomonas from water/biofilm	Pseudomonas → catheter removal usually required; cover with IP ceftazidime/gentamicin, then per sensitivities
Fungal (Candida)	2-5%	Heavily antibiotic-exposed, immunosuppressed, recent bacterial peritonitis	Catheter removal MANDATORY; systemic antifungal (echinocandin/fluconazole); mortality 10-25%
Culture-negative	10-20%	—	Usually responds to IP antibiotics; rule out fastidious organisms (mycobacteria, fungi) if persistent

[1]

Indications for Tenckhoff catheter removal in PD peritonitis

Indication	Reason
Fungal peritonitis (any Candida)	Catheter universally infected/biofilm-coated; cure impossible with catheter in situ. Remove immediately on diagnosis.
Refractory peritonitis (no clinical/culture response at 5 days of appropriate IP antibiotics)	Biofilm; needs catheter removal + 2-3 weeks of systemic antibiotics before reinsertion
Recurrent peritonitis (same organism <30 days after completion of therapy)	Suggests biofilm or tunnel focus; remove catheter
Pseudomonas / Stenotrophomonas	Often refractory; catheter removal usually required to cure
Repeating peritonitis (different organism <30 days after therapy)	Often dialysate/tubing contamination; review exchange technique
Tunnel infection not responding to antibiotics	Suggests catheter focus; removal prevents recurrent
Severe intra-abdominal pathology (appendicitis, ischaemia, perforation)	Differentiate from PD peritonitis (PD = monomicrobial usually); treat as secondary peritonitis + remove catheter

[1]

Empiric management of suspected PD peritonitis — intraperitoneal antibiotics

Diagnosis — cloudy dialysate + abdominal pain ± fever. The cardinal sign is cloudy effluent. Send: dialysate cell count + differential (after ≥2 h dwell), Gram stain, culture. PD peritonitis = dialysate WCC >100/mm³ with PMN >50% (after a dwell).[6]
Empiric IP antibiotics within 2-3 h of presentation — IP vancomycin or cefazolin (gram-positive cover) + IP ceftazidime or gentamicin (gram-negative cover). IP dosing delivers the antibiotic directly to the peritoneum — higher local levels than IV. Loading dose added to one exchange; dwell ≥6 h.
Add heparin 500-1000 U/L to dialysate — prevents fibrin clots obstructing the catheter.
Once culture results (24-72 h) — narrow to organism. CoNS → IP cefazolin or vancomycin alone; S. aureus → IP vancomycin + review exit site; gram-negative → IP ceftazidime (or gentamicin if not aminoglycoside-sparing); fungal → remove catheter + systemic antifungal.
Duration: 2 weeks (coNS, strep); 3 weeks (S. aureus, gram-negative, Enterococcus); until catheter removed + 2-3 weeks systemic (fungal).
Catheter removal if: fungal, refractory at 5 days, recurrent (same organism), Pseudomonas/Stenotrophomonas, or severe intra-abdominal pathology.
Prevention — meticulous exit-site care (mupirocin daily to nares + exit site), training/re-training on aseptic technique (disconnect systems reduce touch contamination), prophylactic mupirocin/rifampicin for S. aureus nasal carriage.

Routes of antibiotic administration in PD peritonitis — IP vs IV

Route	Pros	Cons	When
Intraperitoneal (IP)	Preferred — achieves high peritoneal levels; outperforms IV for PD peritonitis; avoids systemic exposure	Adds to dialysate; requires dwell ≥6 h; compatibility (vancomycin + ceftazidime may precipitate — give in separate exchanges)	First-line for all PD peritonitis; load in the long (overnight) dwell.[6]
Intravenous (IV)	Easy, well tolerated	Lower peritoneal penetration than IP; less effective	Reserve for severe sepsis, systemic infection, or when PD temporarily held; combine with IP

Additional clinical pearls

SBP — diagnosis, treatment and prevention pearls

Diagnostic paracentesis is the single most important investigation in a cirrhotic with ascites and ANY decompensation. Up to a third of SBP cases are culture-negative — PMN ≥ 250 cells/mm³ is the diagnostic threshold, NOT culture positivity. Treating SBP does not require a positive culture.[1] }
Bedside inoculation of ascites into blood-culture bottles doubles yield to 70-80%. Send 10 mL into each bottle at the bedside — the lab-delayed inoculated swab misses half of all SBP.[1] }
Reagent strip (leukocyte esterase dipstick) PMN surrogate is a fast bedside test when lab turnaround is slow — a ≥2+ reading has ~90% sensitivity for SBP.[1] }
Cirrhotic coagulopathy is NOT a contraindication to paracentesis. Mild bleeding <1%, life-threatening bleeding <0.1%. Do NOT give prophylactic FFP/platelets before a tap (these do not prevent the rare visceral injury bleeds anyway).[1] }
Coagulopathy and thrombocytopenia in cirrhosis reflect rebalanced haemostasis — the INR is a poor predictor of bleeding in cirrhosis (pro- and anti-coagulant factors fall in parallel). DO NOT let an abnormal INR delay a diagnostic tap.[1] }
The 48-hour re-tap is essential — repeat paracentesis. PMN should fall by ≥25% at 48 h on appropriate antibiotic. Failure to fall → resistant organism, missed secondary peritonitis, or inadequate source control. Widen empiric cover and image.[1] }
Five days is as effective as ten for uncomplicated SBP — short course reduces resistance and C. difficile. Extend to 7 days only if sepsis, bacteraemia, or delayed response.[1] }
Albumin works by FOUR mechanisms — oncotic (effective circulating volume), anti-inflammatory (binds TNF/endotoxin), renal perfusion (prevents HRS), and reduces bacterial translocation. Sort 1999: HRS 33%→10%, mortality 29%→10%. Albumin dose: 1.5 g/kg day 1, 1 g/kg day 3.[2] }
SBP prophylaxis indications — (a) Previous SBP (lifelong until transplant — recurrence 70% at 1 year without prophylaxis vs 20% with). (b) GI bleed in cirrhosis (7-day course of norfloxacin/ciprofloxacin + ceftriaxone in severe cirrhosis — the ONLY primary prophylaxis with RCT evidence). (c) Low ascitic protein (<15 g/L) WITH impaired renal/liver function (creatinine ≥106 μmol/L, BUN ≥25, sodium ≤130, Child-Pugh ≥9 + bili ≥22 — primary prophylaxis by norfloxacin). (d) All cirrhotic inpatients with ascites — broad-screen on admission is NOT routine, but a low threshold to start prophylaxis after recovery from a first SBP is.[1] }
Quinolone-resistance is rising — long-term norfloxacin prophylaxis selects resistant gut flora. Up to 50% of breakthrough SBP on norfloxacin is gram-positive or quinolone-resistant gram-negative. NEVER use oral ofloxacin/ciprofloxacin empirically for quinolone-prophylaxis SBP — use cefotaxime/ceftriaxone.[3] }
SBP doubles 1-year mortality (30-50%) even with optimal treatment. SBP is a decompensating event — REFER FOR LIVER TRANSPLANT EVALUATION after the first episode. MELD-Na drives listing.[1] }
Bacteraemia occurs in 30-50% of SBP — send blood cultures with the paracentesis. Bacteraemic SBP has worse outcomes (higher HRS risk, mortality). Albumin is even more important in this group.[2] }

Secondary and tertiary peritonitis pearls

Source control is MANDATORY for secondary peritonitis — antibiotics alone cannot cure it. The intra-abdominal source (perforation, leak, abscess, necrosis) must be addressed by drainage, debridement, repair, or resection. Antibiotic penetration into pus/necrotic tissue is poor.[3] }
Surviving Sepsis hour-1 bundle: lactate, blood cultures, antibiotics, 30 mL/kg crystalloid, vasopressors to MAP ≥65. Source control within 6 h of septic shock — each hour of delay in the hypotensive patient increases mortality by ~7-8%.[11] }
Piperacillin/tazobactam is the workhorse empiric regimen for severe community-acquired or healthcare-associated intra-abdominal infection — covers Enterococcus, ESBL GNR, anaerobes, and most Pseudomonas. Switch to meropenem if prior carbapenem-sparing failure, ESBL colonisation, or severe sepsis post-major surgery.[7] }
Cover Enterococcus in healthcare-associated / postoperative IAI — E. faecium is selected by cephalosporins and is the dominant Enterococcus in nosocomial infection; ~50% are VRE in some ICUs. Add vancomycin empirically in post-op/nosocomial IAIs.[7] }
Add an echinocandin (caspofungin) for Candida peritonitis risk factors — recent fluoroquinolone/cephalosporin exposure, TPN, multiple laparotomies, necrotising pancreatitis, immunosuppression, prolonged ICU stay. Switch to fluconazole if stable and C. albicans susceptible.[10] }
STOP-IT: 4 days of antibiotics after adequate source control — non-inferior to ~8 days. Prolonged courses drive resistance and C. difficile without benefit. Continue beyond 4 days ONLY if: bacteraemia/fungaemia, persistent sepsis, incomplete source control, or immunosuppression.[4] }
Use WATER-SOLUBLE contrast (gastrograffin/omnipaque) — NEVER barium — for suspected anastomotic leak or perforation. Barium in the peritoneum causes severe granulomatous peritonitis and is catastrophic if surgery follows.[11] }
Damage-control surgery for the unstable patient: temporary abdominal closure (VAC / Bogota bag), planned second-look laparotomy in 24-48 h. Definitive anastomosis is deferred until physiology restored — primary anastomosis in an unstable, acidotic, coagulopathic patient will leak.[11] }
Tertiary peritonitis is a disease of the failing host, not uncontrolled infection. Cultures are often MDR/low-virulence organisms (Enterococcus, Candida, Pseudomonas, Stenotrophomonas). Surgery often finds "sticky" peritoneum without a clear source — the operation is unrewarding. Focus on broad anti-MDR cover, antifungal, enteral nutrition, and supportive ICU care.[5] }
Stenotrophomonas in tertiary peritonitis → think carbapenem exposure. The only reliable agent is trimethoprim-sulfamethoxazole. A Stenotrophomonas isolate in abdominal cultures tells you the patient has had prolonged carbapenem exposure and immune paresis.[5] }
Tertiary peritonitis may be culture-NEGATIVE despite florid sepsis — the host is too immunoparalysed to mount a neutrophil response, and antibiotics have sterilised cultures. Persistent leucopenia, thrombocytopenia, rising lactate, vasopressor dependence are markers. Treat empirically with broad-spectrum + antifungal.[5] }
Candida peritonitis is a marker of severe critical illness, not the cause. Mortality 40-60% even with appropriate antifungal therapy. Source control (remove infected material, drain collections) and host support matter as much as the antifungal.[10] }
The immunocompromised host (transplant, chemotherapy, neutropenia, chronic steroids) presents atypically — often afebrile, blunted abdominal signs, late presentation. Lower threshold for CT and surgical source control. WSES 2021 immunocompromised guidelines recommend early imaging and broad empiric cover.[12] }
Intra-abdominal hypertension (IAH) and abdominal compartment syndrome (ACS) complicate severe peritonitis — measure intravesical pressure; IAP >12 mmHg = IAH, >20 mmHg with new organ failure = ACS. Surgical decompression (open abdomen) may be required to relieve ACS, prevent renal failure, and improve ventilation.[11] }
Procalcitonin trends can guide antibiotic duration in intra-abdominal infection — falling levels support early stopping at 4 days. A persistently elevated procalcitonin suggests ongoing infection, undrained source, or tertiary peritonitis.[11] }

Catheter-related (PD) peritonitis and special situations

Cloudy dialysate + abdominal pain = PD peritonitis until proven otherwise. Send the FIRST cloudy bag for cell count + culture. WCC >100/mm³ with PMN >50% (after ≥2 h dwell) confirms PD peritonitis.[6] }
Intraperitoneal (IP) antibiotics outperform IV in PD peritonitis. IP vancomycin/cefazolin (gram-positive) + IP ceftazidime/gentamicin (gram-negative) as initial empiric therapy, loaded in the long (overnight) dwell. Vancomycin + ceftazidime may precipitate — give in separate exchanges.[6] }
Add heparin 500-1000 U/L to dialysate — prevents fibrin clots that obstruct the catheter and reduce antibiotic delivery during peritonitis. }
Coagulase-negative staphylococcal PD peritonitis = touch contamination. Mild course, usually responds to IP cefazolin/vancomycin, catheter retained. Review exchange technique (the patient re-touches the disconnect system with imperfect asepsis).[6] }
Staphylococcus aureus PD peritonitis — examine the exit site and tunnel. S. aureus often comes from an exit-site/tunnel infection; cure is unlikely without addressing the catheter focus. Nasal mupirocin reduces S. aureus carriage and exit-site infection.[6] }
Pseudomonas / Stenotrophomonas PD peritonitis usually needs catheter removal. Biofilm-coated catheters are impossible to sterilise with antibiotics alone. Combine IP ceftazidime/gentamicin with catheter removal + 2-3 weeks of systemic antibiotics.[6] }
Fungal PD peritonitis → remove the catheter IMMEDIATELY on diagnosis. Mortality 10-25%; cure impossible with catheter in situ. Start systemic echinocandin → fluconazole step-down. Risk factors: recent bacterial peritonitis (prolonged antibiotics), immunosuppression, diabetes.[6] }
Refractory PD peritonitis (no clinical/culture response at 5 days) → catheter removal. Biofilm-coated catheter, infection with Pseudomonas/fungal, or tunnel infection. Continuing IP antibiotics beyond 5 days without response risks systemic sepsis and peritoneal damage.[6] }
PD peritonitis is monomicrobial in 80% of cases — polymicrobial PD peritonitis (especially with anaerobes) suggests an intra-abdominal source (perforation, ischaemia) — treat as secondary peritonitis and image. Catheter removal often required.[6] }
Antibiotic prophylaxis before PD catheter placement — single dose of IV cefazolin at induction reduces early PD peritonitis. Exit-site mupirocin daily reduces S. aureus exit-site infection and peritonitis.[6] }

Additional red flags

Critical peritonitis red flags — the warning features

SBP with rising creatinine or bilirubin >68 μmol/L → give albumin 1.5 g/kg day 1, 1 g/kg day 3 (non-negotiable; prevents HRS, halves mortality).[2] }
SBP not responding at 48 h (PMN fall <25%) → resistant organism or missed secondary peritonitis — broaden to meropenem, repeat the tap, image for perforation.[1] }
Polymicrobial ascitic culture or anaerobes → this is NOT SBP. Suspect perforation → CT + surgery.[1] }
Quinolone-prophylaxis patient with SBP → do NOT use oral ofloxacin/ciprofloxacin. Use cefotaxime/ceftriaxone — high quinolone resistance.[3] }
Tertiary peritonitis → broad-spectrum anti-MDR + echinocandin, address immune paresis, mortality 30-50%.[5] }
Stenotrophomonas in tertiary peritonitis → only reliable agent is trimethoprim-sulfamethoxazole; selected by carbapenem exposure.[5] }
Candida peritonitis in critically ill / postoperative → echinocandin (caspofungin), source control, mortality 40-60%.[10] }
Septic shock from abdominal source → Surviving Sepsis hour-1 bundle; source control within 6 h.[11] }
Anastomotic leak post-surgery (day 3-7) → fever, ileus, drain output (faeculent/bilious) → re-operation; image with water-soluble contrast NEVER barium.[3] }
Abdominal compartment syndrome (intra-abdominal pressure >20 mmHg with new organ failure) → surgical decompression, open abdomen.[11] }
PD peritonitis with fungal / Pseudomonas / refractory at 5 days / recurrent same organism → catheter removal MANDATORY.[6] }
Polymicrobial PD peritonitis (with anaerobes) → intra-abdominal source (perforation), not PD peritonitis — treat as secondary peritonitis.[6] }
Cirrhotic with "valentino's sign" (RIF pain from gastric fluid tracking down the right paracolic gutter) → perforated PUD, erect CXR for free gas.[3] }
Immunosuppressed patient (transplant, chemo, neutropenia) with abdominal pain → atypical, often afebrile; early CT + surgical source control; broad empiric cover per WSES 2021.[12] }

Summary — what the CICM/FFICM/EDIC candidate must know

The 10 facts that come up in the exam

SBP diagnosis = ascitic PMN ≥ 250 cells/mm³ — culture often negative, do not wait.
SBP treatment = cefotaxime/ceftriaxone + ALBUMIN (1.5 g/kg D1, 1 g/kg D3) — albumin halves HRS and mortality.
SBP organisms: monomicrobial — E. coli, Klebsiella, S. pneumoniae (anaerobes rare).
Secondary peritonitis = polymicrobial + surgical source control MANDATORY; antibiotics alone insufficient.
STOP-IT: 4 days of antibiotics after adequate source control (no benefit of longer).
Tertiary peritonitis = MDR organisms (Enterococcus, Candida, Pseudomonas) + 30-50% mortality; host immune failure.
CAPD peritonitis = cloudy dialysate, WCC >100/mm³ with PMN >50%; intraperitoneal antibiotics first-line.
Catheter removal in PD peritonitis: fungal, refractory, recurrent, Pseudomonas.
Surviving Sepsis hour-1 bundle + source control within 6 h for septic shock from abdominal source.
Albumin trials: Sort 1999 (SBP, benefit), ANSWER 2018 (long-term outpatient, benefit), ATTIRE 2021 (inpatient target, no benefit) — be selective.

[1]

Peritonitis at a glance — the one-table revision

Feature	SBP	Secondary	Tertiary	CAPD
Population	Cirrhosis + ascites	Any abdominal catastrophe	Critically ill ICU patient	PD patient
Source	None (haematogenous)	Perforation/leak/abscess	Persistent/recurrent; often no source	Catheter (touch contamination)
Microbiology	Monomicrobial GNR (E. coli, Klebsiella, S. pneumo)	Polymicrobial gut flora + anaerobes	MDR (Enterococcus, Candida, Pseudomonas, MRSA)	Staph (CoNS, S. aureus), GNR, fungal
Diagnosis	Ascitic PMN ≥ 250/mm³	CT + surgery	Persistent sepsis, cultures positive	Dialysate WCC >100/mm³, PMN >50%
First-line treatment	Cefotaxime + ALBUMIN	Surgery + piperacillin/tazobactam	Meropenem + vancomycin + echinocandin	Intraperitoneal vancomycin/cefazolin + ceftazidime
Source control	NOT required (no source)	MANDATORY	Often unrewarding	Catheter removal for specific indications
Antibiotic duration	5-7 days	4 days after source control	Often prolonged (until clinically resolving)	2-3 weeks
Mortality	10-30% (albumin halves high-risk)	5-20% (driven by source-control timing)	30-50%	<5% per episode
Key trial	Sort 1999 (albumin)	STOP-IT 2015 (4 days)	Evans 2001 (tertiary phenotype)	ISPD 2016 (IP antibiotics)

[1]

References

[1]Runyon BA, AASLD. Introduction to the revised American Association for the Study of Liver Diseases Practice Guideline management of adult patients with ascites due to cirrhosis 2012 Hepatology, 2013.PMID 23463403
[2]Sort P, Navasa M, Arroyo V, et al. Effect of intravenous albumin on renal impairment and mortality in patients with cirrhosis and spontaneous bacterial peritonitis N Engl J Med, 1999.PMID 10432325
[3]Solomkin JS, Mazuski JE, Bradley JS, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America Clin Infect Dis, 2010.PMID 20034345
[4]Sawyer RG, Claridge JA, Nathens AB, et al. Trial of short-course antimicrobial therapy for intraabdominal infection N Engl J Med, 2015.PMID 25992746
[5]Evans HL, Raymond DP, Pelletier SJ, et al. Tertiary peritonitis (recurrent diffuse or localized disease) is not an independent predictor of mortality in surgical patients with intraabdominal infection Surg Infect (Larchmt), 2001.PMID 12593701
[6]Li PK-T, Szeto CC, Piraino B, et al. ISPD Peritonitis Recommendations: 2016 Update on Prevention and Treatment Perit Dial Int, 2016.PMID 27282851
[7]Mazuski JE, Tessier JM, May AK, et al. The Surgical Infection Society Revised Guidelines on the Management of Intra-Abdominal Infection Surg Infect (Larchmt), 2017.PMID 28085573
[8]Caraceni P, Riggio O, Angeli P, et al.; ANSWER Study Investigators. Long-term albumin administration in decompensated cirrhosis (ANSWER): an open-label randomised trial Lancet, 2018.PMID 29861076
[9]China L, Freemantle N, Forrest E, et al.; ATTIRE trial investigators. A Randomized Trial of Albumin Infusions in Hospitalized Patients with Cirrhosis N Engl J Med, 2021.PMID 33657293
[10]Pappas PG, Kauffman CA, Andes D, et al. Clinical practice guidelines for the management of candidiasis: 2009 update by the Infectious Diseases Society of America Clin Infect Dis, 2009.PMID 19191635
[11]Sartelli M, Mazuski JE, Coccolini F, et al. Intra-abdominal infections survival guide: a position statement by the Global Alliance For Infections In Surgery World J Emerg Surg, 2024.PMID 38851700
[12]Coccolini F, Sartelli M, Podda M, et al. Acute abdomen in the immunocompromised patient: WSES, SIS-E, WSIS, AAST, and GAIS guidelines World J Emerg Surg, 2021.PMID 34372902

Your progress

SAQ — Spontaneous bacterial peritonitis in a cirrhotic with decompensation

SAQ — Secondary peritonitis from a perforated viscus with septic shock

Clinical pearls

Red flags

Prognosis

Albumin in SBP (Sort 1999, NEJM)

SBP — spontaneous bacterial peritonitis: deep dive [1]

ANSWER trial — Caraceni 2018, Lancet (long-term albumin in decompensated cirrhosis)

ATTIRE — China 2021, NEJM (albumin in hospitalised cirrhosis)

Secondary peritonitis: deep dive — sources, organisms, source control

STOP-IT trial — Sawyer 2015, NEJM (duration of antibiotics after source control)

Tertiary peritonitis: deep dive — the failing host

Additional clinical pearls

Additional red flags

Summary — what the CICM/FFICM/EDIC candidate must know

References

Peritonitis in ICU: spontaneous bacterial, secondary, and tertiary

Your progress

SAQ — Spontaneous bacterial peritonitis in a cirrhotic with decompensation

SAQ — Secondary peritonitis from a perforated viscus with septic shock

Clinical pearls

Red flags

Prognosis

Albumin in SBP (Sort 1999, NEJM)

SBP — spontaneous bacterial peritonitis: deep dive [1]

ANSWER trial — Caraceni 2018, Lancet (long-term albumin in decompensated cirrhosis)

ATTIRE — China 2021, NEJM (albumin in hospitalised cirrhosis)

Secondary peritonitis: deep dive — sources, organisms, source control

STOP-IT trial — Sawyer 2015, NEJM (duration of antibiotics after source control)

Tertiary peritonitis: deep dive — the failing host

Additional clinical pearls

Additional red flags

Summary — what the CICM/FFICM/EDIC candidate must know

References

Your progress

SAQ — Spontaneous bacterial peritonitis in a cirrhotic with decompensation

SAQ — Secondary peritonitis from a perforated viscus with septic shock

Clinical pearls

Red flags

Prognosis

Albumin in SBP (Sort 1999, NEJM)

SBP — spontaneous bacterial peritonitis: deep dive [1]

ANSWER trial — Caraceni 2018, Lancet (long-term albumin in decompensated cirrhosis)

ATTIRE — China 2021, NEJM (albumin in hospitalised cirrhosis)

Secondary peritonitis: deep dive — sources, organisms, source control

STOP-IT trial — Sawyer 2015, NEJM (duration of antibiotics after source control)

Tertiary peritonitis: deep dive — the failing host

Catheter-related / peritoneal dialysis (CAPD) peritonitis

Additional clinical pearls

Additional red flags

Summary — what the CICM/FFICM/EDIC candidate must know

References

Your progress

SAQ — Spontaneous bacterial peritonitis in a cirrhotic with decompensation

SAQ — Secondary peritonitis from a perforated viscus with septic shock

Clinical pearls

Red flags

Prognosis

Albumin in SBP (Sort 1999, NEJM)

SBP — spontaneous bacterial peritonitis: deep dive [1]

ANSWER trial — Caraceni 2018, Lancet (long-term albumin in decompensated cirrhosis)

ATTIRE — China 2021, NEJM (albumin in hospitalised cirrhosis)

Secondary peritonitis: deep dive — sources, organisms, source control

STOP-IT trial — Sawyer 2015, NEJM (duration of antibiotics after source control)

Tertiary peritonitis: deep dive — the failing host

Catheter-related / peritoneal dialysis (CAPD) peritonitis

Additional clinical pearls

Additional red flags

Summary — what the CICM/FFICM/EDIC candidate must know

References